classification/ClassificationMulti2Features-Tree.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Multiclass classification with 2 features using trees\n",
    "\n",
    "Classification trees are a wide category of algorithms to process classification (and linear regression).\n",
    "\n",
    "In this notebook, we will reuse the generated dataset of the multiclass linear regression with Keras ([HTML](ClassificationMulti2Features-Keras.html) / [Jupyter](ClassificationMulti2Features-Keras.ipynb)) : the Czech and Norvegian flags.\n",
    "\n",
    "Learning goals:\n",
    "- Theory of trees\n",
    "- Custom code for trees\n",
    "- Understanding of the tree partitioning and performance\n",
    "- Decision tree using Scikit-Learn model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.colors as pltcolors\n",
    "from sklearn import metrics, tree\n",
    "import graphviz\n",
    "from IPython.display import SVG\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Helpers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "def flagPlot(X, y, title, colors, ax=None):\n",
    "    \"\"\" Plot the flag as a 2D parametric label\"\"\"\n",
    "    if ax is None:\n",
    "        ax = plt.gca()\n",
    "    ax.set_facecolor((0.8, 0.8, 0.8))\n",
    "    f = ax.scatter(X[:,0], X[:,1], c=y, cmap=pltcolors.ListedColormap(colors), marker='x', alpha=0.5);\n",
    "    ax.set_xlabel('$x_0$')\n",
    "    ax.set_ylabel('$x_1$')\n",
    "    ax.grid()\n",
    "    ax.set_title(title)\n",
    "    cb = plt.colorbar(f, ax=ax)\n",
    "    loc = np.arange(0,2.1,1)\n",
    "    cb.set_ticks(loc)\n",
    "    cb.set_ticklabels([0,1,2]);\n",
    "    \n",
    "def plotHeatMap(X, classes, title=None, fmt='.2g', ax=None, xlabel=None, ylabel=None):\n",
    "    \"\"\" Fix heatmap plot from Seaborn with pyplot 3.1.0, 3.1.1\n",
    "        https://stackoverflow.com/questions/56942670/matplotlib-seaborn-first-and-last-row-cut-in-half-of-heatmap-plot\n",
    "    \"\"\"\n",
    "    ax = sns.heatmap(X, xticklabels=classes, yticklabels=classes, annot=True, \\\n",
    "                     fmt=fmt, cmap=plt.cm.Blues, ax=ax) #notation: \"annot\" not \"annote\"\n",
    "    bottom, top = ax.get_ylim()\n",
    "    ax.set_ylim(bottom + 0.5, top - 0.5)\n",
    "    if title:\n",
    "        ax.set_title(title)\n",
    "    if xlabel:\n",
    "        ax.set_xlabel(xlabel)\n",
    "    if ylabel:\n",
    "        ax.set_ylabel(ylabel)\n",
    "        \n",
    "def plotConfusionMatrix(yTrue, yEst, classes, title=None, fmt='.2g', ax=None):\n",
    "    plotHeatMap(metrics.confusion_matrix(yTrue, yEst), classes, title, fmt, ax, xlabel='Estimations', \\\n",
    "                ylabel='True values');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data models - Flags"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "nFeatures = 2\n",
    "nClasses = 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Czech flag colors\n",
    "czechColors = np.array(['blue', 'red', 'white'])\n",
    "def generateCzechBatch(n, noiseSigma = 0.1):\n",
    "    \"\"\" Generate a multi class sample based on the Czech banner geometry. \n",
    "        Tip of the triangle is set at the origin \n",
    "    \"\"\"\n",
    "    # xMax adjusted such that the 3 classes are quasi equiprobable\n",
    "    xMin = np.array([-1, -1])\n",
    "    xMax = np.array([0.5, 1])\n",
    "    #\n",
    "    X = np.random.uniform(xMin, xMax, (n, nFeatures))\n",
    "    noisyX = X + np.random.normal(0, noiseSigma, X.shape)\n",
    "    y = np.zeros(n)\n",
    "    y[noisyX[:,1] > 0]  = 2                                       # White\n",
    "    y[noisyX[:,1] <= 0] = 1                                       # Red\n",
    "    y[(noisyX[:,0] <= 0) & (np.abs(noisyX[:,1]) < np.abs(noisyX[:,0]))] = 0 # Blue triangle\n",
    "    return X, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Norway flag colors\n",
    "norwayColors = ['red', 'white', 'navy']\n",
    "def generateNorwayBatch(n, noiseSigma = 0.1):\n",
    "    \"\"\" Generate a multi class sample based on the Norway banner geometry \"\"\"\n",
    "    # xMax adjusted such that the 3 classes are quasi equiprobable\n",
    "    xMin = np.array([-1, -1])\n",
    "    xMax = np.array([1, 1])\n",
    "    #\n",
    "    X = np.random.uniform(xMin, xMax, (n, nFeatures))\n",
    "    a1 = 0.43\n",
    "    a2 = 0.18\n",
    "    noisyX = X + np.random.normal(0, noiseSigma, X.shape)\n",
    "    y = np.zeros(n)                                                                # Red = background\n",
    "    y[((noisyX[:,0] > -a1) & (noisyX[:,0] < a1)) | ((noisyX[:,1] > -a1) & (noisyX[:,1] < a1))]  = 1    # White cross\n",
    "    y[((noisyX[:,0] > -a2) & (noisyX[:,0] < a2)) | ((noisyX[:,1] > -a2) & (noisyX[:,1] < a2))]  = 2    # Navy cross over white\n",
    "    return X, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x288 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "N = 1000\n",
    "xTrainC, yTrainC = generateCzechBatch(N)\n",
    "fig, axes = plt.subplots(1, 2, figsize=(15,4))\n",
    "flagPlot(xTrainC, yTrainC, 'Generated Czech flag', czechColors, axes[0])\n",
    "xTrainN, yTrainN = generateNorwayBatch(N)\n",
    "flagPlot(xTrainN, yTrainN, 'Generated Norway flag', norwayColors, axes[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "xTestC, yTestC = generateCzechBatch(N)\n",
    "xTestN, yTestN = generateNorwayBatch(N)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tree classifier\n",
    "\n",
    "Tree classifiers are a very large family of non-parametric classifiers, and have many refinements. They are splitting recursively a portion of the space with an hyperplan. \n",
    "\n",
    "In our case, the space is 2 dimension, the splitting hyperplan is thus 1D : a line. \n",
    "\n",
    "We will first implement a tree classifier from scratch, then use the Scikit-Learn implementation.\n",
    "\n",
    "Trees main advantages are :\n",
    "- Better ability to approximate complex functions than linear regression\n",
    "- Explanability of the decisions since the boundaries are known. However, the decisions might have no human understanding\n",
    "- Far more efficient than the kNN classifier ([HTML](ClassificationContinuous2Features-KNN.html) / [Jupyter](ClassificationContinuous2Features-KNN.ipynb)) since the model is fitted only once\n",
    "\n",
    "Their main drawbacks : \n",
    "- Cost and difficulty to reach the optimum. We will see later the Random Forest method that is mitigating this issue"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Algorithm\n",
    "\n",
    "We will use a simple recursive split algorithm similar to CART. \n",
    "\n",
    "At each iteration of the split :\n",
    "- Select the feature (co-variable) bringing the best improvement on the impurity (also known as information gain)\n",
    "- Split the space with an hyperplan orthogonal to that feature's axis\n",
    "\n",
    "For simplification purpose, following algo choices are set:\n",
    "- Impurity improvement is measured as entropy minimization, other metrics are often in use like Gini\n",
    "- Split is at the middle point of the axis, finer grain split is possible\n",
    "- Partitioning stop condition is on the leaf minimum item number and the minimum entropy\n",
    "\n",
    "Impurity is calculated as the average of the entropies of the two partitions created by the split, weighted by the number of items in each partition.\n",
    "\n",
    "Entropy is defined using the empirical probabilities of the $Y$ labels within each classes : \n",
    "$$ - \\sum_{c \\in C} \\mathbb{P}(y=c) log(\\mathbb{P}(y=c))$$\n",
    "\n",
    "$C$ is the set of classes in $Y$, in our case $ C = \\{0, 1, 2\\}$ or the set of three colors.\n",
    "\n",
    "\n",
    "### References\n",
    "\n",
    "- [Classification and regression trees (CART), Leo Breiman, Joseph H Friedman, C. J. Stone - 1984](https://doi.org/10.2307/2530946)\n",
    "- [Decision Tree tutorial at Analytics Vidhya](https://www.analyticsvidhya.com/blog/2016/04/complete-tutorial-tree-based-modeling-scratch-in-python/#three)\n",
    "- [Let’s Write a Decision Tree Classifier from Scratch - Machine Learning Recipes #8 - Youtube](https://www.youtube.com/watch?v=LDRbO9a6XPU)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Home made implementation\n",
    "\n",
    "During training, the classifier algorithm is partitioning the original X space and producing a tree representing this partitioning.\n",
    "\n",
    "Two Python classes are designed :\n",
    "- The classifier _MyTreeClassifier_\n",
    "- The tree element _TreeNode_ that may be intermediate or terminal. Intermediate nodes have two children\n",
    "\n",
    "The algorithm used for the partitioning is quite naive as it heavily uses numpy slicing based on range inequalities. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "class TreeNode:\n",
    "    \"\"\" Tree node element\n",
    "        Either an intermediate node with two children (left-right) or a leaf node\n",
    "    \"\"\"\n",
    "    \n",
    "    def __init__(self, depth, parent, boundary, entropy):\n",
    "        self.depth = depth\n",
    "        self.parent = parent\n",
    "        self.boundary = boundary\n",
    "        self.splitOnFeature = None\n",
    "        self.halfPoint = None\n",
    "        self.entropy = entropy\n",
    "        self.leftChild = None\n",
    "        self.rightChild = None\n",
    "        self.valueFrequencies = None\n",
    "        self.isLeaf = False\n",
    "        \n",
    "    def split(self, splitOnFeature, halfPoint, childL, childR, valueFrequencies=None):\n",
    "        \"\"\" Split node into to two child nodes \"\"\"\n",
    "        self.splitOnFeature = splitOnFeature\n",
    "        self.halfPoint = halfPoint\n",
    "        self.leftChild = childL\n",
    "        self.rightChild = childR\n",
    "        self.valueFrequencies = valueFrequencies\n",
    "   \n",
    "    def setAsLeaf(self, valueFrequencies):\n",
    "        \"\"\" Set this node to be a terminal (a leaf) \"\"\"\n",
    "        self.valueFrequencies = valueFrequencies\n",
    "        self.isLeaf = True\n",
    "        \n",
    "    def predict(self, x):\n",
    "        \"\"\" Recursively predict value \"\"\"\n",
    "        if len(x.shape) > 1:\n",
    "            if self.isLeaf:\n",
    "                return self.getValue() * np.ones(x.shape[0])\n",
    "            else:\n",
    "                result = np.zeros(x.shape[0])\n",
    "                left = (x[:,self.splitOnFeature] <= self.halfPoint)\n",
    "                result[left] = self.leftChild.predict(x[left])\n",
    "                result[~left] = self.rightChild.predict(x[~left])\n",
    "                return result\n",
    "        else:\n",
    "            if self.isLeaf:\n",
    "                return self.getValue()\n",
    "            elif(x[self.splitOnFeature] > self.halfPoint):\n",
    "                return self.rightChild.predict(x)\n",
    "            else:\n",
    "                return self.leftChild.predict(x)\n",
    "\n",
    "    def getValue(self):\n",
    "        \"\"\" Take the class with highest frequency as value \"\"\"\n",
    "        return np.argmax(self.valueFrequencies)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "class MyTreeClassifier:\n",
    "    \"\"\" Tree classifier using CART recursive algorithm \n",
    "        Stop criteria are the leaf minimum number of items, and the minimum entropy\n",
    "        Entropy is used to measure the impurity\n",
    "    \"\"\"\n",
    "\n",
    "    def __init__(self, leafMinSize=5, entropyMin=1e-5):\n",
    "        self.leafMinSize = leafMinSize\n",
    "        self.entropyMin = entropyMin\n",
    "        self.root = None\n",
    "\n",
    "    def fit(self, X, y, yClasses=None):\n",
    "        \"\"\" Partition X partition in order to predict y \"\"\"\n",
    "        if not yClasses:\n",
    "            yClasses = np.unique(y)\n",
    "        p = X.shape[1]\n",
    "        # Stack X and y in order to extract slices\n",
    "        Xy = np.hstack((X, y.reshape(-1, 1)))\n",
    "        # Get overall X bounding box\n",
    "        epsilon = 1e-4\n",
    "        originalBoundary = [[X[:,j].min() - epsilon, X[:,j].max()] for j in range(p)]\n",
    "        \n",
    "        # Initialize FIFO of sub-partitions to split, \n",
    "        #  it avoids recursive calls with large sub-partitions in memory\n",
    "        self.root = TreeNode(0, None, originalBoundary, self.getEntropy(y, yClasses))\n",
    "        splitList = [(Xy, self.root)]\n",
    "        processedList = []\n",
    "\n",
    "        # Run until splitList FIFO empty\n",
    "        while splitList:\n",
    "            partition, node = splitList.pop(0)\n",
    "            selectedFeature, halfPoint, newBoundary, entropies = \\\n",
    "                self.splitDecision(partition, node.boundary, yClasses)\n",
    "            partitionValue = self.getValueFrequencies(partition[:,nFeatures], len(yClasses))\n",
    "            \n",
    "            # Split\n",
    "            children  = [TreeNode(node.depth + 1, node, newBoundary[i], entropies[i]) for i in range(2)]\n",
    "            node.split(selectedFeature, halfPoint, children[0], children[1], partitionValue)\n",
    "            processedList.append(node)\n",
    "            \n",
    "            # (Schedule) process of children\n",
    "            for child, bound, entropy in zip(children, newBoundary, entropies):\n",
    "                subpartition = self.selectSubPartition(partition, bound)\n",
    "                if (len(subpartition) > self.leafMinSize) and (entropy > self.entropyMin):\n",
    "                    splitList.append((subpartition, child))\n",
    "                else:\n",
    "                    child.setAsLeaf(self.getValueFrequencies(subpartition[:,nFeatures], len(yClasses)))\n",
    "                    processedList.append(child)\n",
    "        return processedList\n",
    "      \n",
    "    def predict(self, X):\n",
    "        \"\"\" Predict y values given X \"\"\"\n",
    "        assert(self.root != None)\n",
    "        return self.root.predict(X)   \n",
    "\n",
    "    def splitDecision(self, Xy, xBoundary, yClasses):\n",
    "        \"\"\" Select variable axis to be split computing impurity \"\"\"\n",
    "        p = len(xBoundary)\n",
    "        nTotal = len(Xy)\n",
    "        # Distribution after split\n",
    "        halfPoints = [(xBoundary[i][0] + xBoundary[i][1])/2 for i in range(nFeatures)]\n",
    "        impurity = []\n",
    "        entropies = []\n",
    "        for xIndex in range(nFeatures):\n",
    "            xRangeLeft  = (Xy[:,xIndex] > xBoundary[xIndex][0]) & (Xy[:,xIndex] <= halfPoints[xIndex])\n",
    "            xRangeRight = (Xy[:,xIndex] > halfPoints[xIndex]) & (Xy[:,xIndex] <= xBoundary[xIndex][1])\n",
    "            weightedEntropies = []\n",
    "            for ra in [xRangeLeft, xRangeRight]:\n",
    "                partitionHalf  = Xy[ra]\n",
    "                nPartition  = len(partitionHalf)\n",
    "                if nPartition > 0:\n",
    "                    entropy = self.getEntropy(partitionHalf[:,nFeatures], yClasses)\n",
    "                    entropies.append(entropy)\n",
    "                    weightedEntropies.append(entropy * nPartition)\n",
    "                else:\n",
    "                    entropies.append(0)\n",
    "            impurity.append(np.sum(weightedEntropies) / nTotal)\n",
    "        # Select the feature to split that is leading to the lowest impurity\n",
    "        selectedFeature = np.argmin(impurity)\n",
    "        newBoundary = [[[xBoundary[i][0], halfPoints[i] if i == selectedFeature else xBoundary[i][1]] \\\n",
    "                        for i in range(nFeatures)],\n",
    "                      [[halfPoints[i] if i == selectedFeature else xBoundary[i][0], xBoundary[i][1]] \\\n",
    "                        for i in range(nFeatures)]]\n",
    "        return selectedFeature, halfPoints[selectedFeature], newBoundary, \\\n",
    "            entropies[selectedFeature*2:(selectedFeature+1)*2]\n",
    "    \n",
    "    def getEntropy(self, y, yClasses):\n",
    "        \"\"\" Compute entropy given the empirical probabilities for each class\"\"\"\n",
    "        nPartition = len(y)\n",
    "        frequencies = np.array([np.sum(y == c) / nPartition for c in yClasses])\n",
    "        # epsilon added to avoid log of 0\n",
    "        epsilon = 1e-10\n",
    "        return - np.sum(np.dot(frequencies, np.log(frequencies + epsilon)))\n",
    "\n",
    "    def selectSubPartition(self, Xy, xBoundary):\n",
    "        \"\"\" Select a rectangular sub-partition \"\"\"\n",
    "        p = len(xBoundary)\n",
    "        xRanges = [(Xy[:,j] > xBoundary[j][0]) & (Xy[:,j] <= xBoundary[j][1]) for j in range(p)]\n",
    "        return Xy[xRanges[0] & xRanges[1]]\n",
    "    \n",
    "    def getValueFrequencies(self, y, nClasses):\n",
    "        binEdges = np.arange(nClasses + 1) - 0.5\n",
    "        return np.histogram(y, bins=binEdges)[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model Fit on Czech and Norway flags"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "model0Czech = MyTreeClassifier()\n",
    "processedListC = model0Czech.fit(xTrainC, yTrainC, range(3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "model0Norway = MyTreeClassifier()\n",
    "processedListN = model0Norway.fit(xTrainN, yTrainN, range(3))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Analysis and evaluation\n",
    "\n",
    "#### Plot tree partitioning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plotNodeArea(node, palette, ax):\n",
    "    \"\"\" Plot the area and split line of a tree node \"\"\"\n",
    "    color = palette[int(np.round(node.getValue()))]\n",
    "    ax.fill_between(node.boundary[0], node.boundary[1][0], node.boundary[1][1], color=color)\n",
    "    if node.halfPoint:\n",
    "        if node.splitOnFeature == 0:\n",
    "            return ax.plot(np.ones(2) * node.halfPoint, node.boundary[1], color='k')\n",
    "        else:\n",
    "            return ax.plot(node.boundary[0], np.ones(2) * node.halfPoint, color='k')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1, 2, figsize=(12, 5))\n",
    "for ax, processedList, palette in zip(axes, [processedListC, processedListN], [czechColors, norwayColors]):\n",
    "    ax.set_facecolor((0.8, 0.8, 0.8))\n",
    "    for targetDepth in range(1, 10):\n",
    "        for n in filter(lambda n: n.depth == targetDepth, processedList):\n",
    "            plotNodeArea(n, palette, ax)\n",
    "\n",
    "axes[0].set_title(\"Czech flag classification tree\")\n",
    "axes[0].set_xlim(-1, 0.5)\n",
    "axes[0].set_ylim(-1, 1);\n",
    "axes[1].set_title(\"Norway flag classification tree\")\n",
    "axes[1].set_xlim(-1, 1)\n",
    "axes[1].set_ylim(-1, 1);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Czech flag is quite homogeneous within the three color areas. Challenges are the color area boundaries where the gaussian noise is mixing color points, and the diagonal boundaries that mus be approximated by small rectangles.\n",
    "\n",
    "The Norway flag is more complex because of the small width of the white area. Noise is also adding some challenges. We observe that the partitioning is not similar on the four quadrants."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Weighted entropy plot for Norway flag\n",
    "\n",
    "All along the tree fit, the weighted entropy is computed for each split area (left / right)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def getMinMaxEntropies(nodeList, depthRange):\n",
    "    weithedEntropiesMin = []\n",
    "    weithedEntropiesMax = []\n",
    "    for targetDepth in depthRange:\n",
    "        weithedEntropies = np.array([node.entropy * np.sum(node.valueFrequencies) for node in filter(lambda n: n.depth == targetDepth, nodeList)])\n",
    "        weithedEntropiesMin.append(weithedEntropies.min())\n",
    "        weithedEntropiesMax.append(weithedEntropies.max())\n",
    "    return weithedEntropiesMin, weithedEntropiesMax\n",
    "\n",
    "depthRange = range(1, 10)\n",
    "entropiesMin, entropiesMax = getMinMaxEntropies(processedListN, depthRange)\n",
    "fig, ax = plt.subplots(1, 1, figsize=(12, 5))\n",
    "ax.fill_between(depthRange, entropiesMin, entropiesMax)\n",
    "ax.set_xlabel(\"Depth\")\n",
    "ax.set_ylabel(\"Weighted entropy\")\n",
    "ax.set_title(\"Min and max entropy weighted by partition size - Norway flag\")\n",
    "ax.grid()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Czech flag classifier performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.86      0.89      0.88       336\n",
      "         1.0       0.93      0.91      0.92       328\n",
      "         2.0       0.92      0.90      0.91       336\n",
      "\n",
      "    accuracy                           0.90      1000\n",
      "   macro avg       0.90      0.90      0.90      1000\n",
      "weighted avg       0.90      0.90      0.90      1000\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "yEstC = model0Czech.predict(xTestC)\n",
    "plotConfusionMatrix(yTestC, yEstC, czechColors, title=\"Czech flag confusion matrix\")\n",
    "print(metrics.classification_report(yTestC, yEstC))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Norway flag classifier performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.80      0.84      0.82       319\n",
      "         1.0       0.59      0.62      0.61       366\n",
      "         2.0       0.71      0.64      0.67       315\n",
      "\n",
      "    accuracy                           0.70      1000\n",
      "   macro avg       0.70      0.70      0.70      1000\n",
      "weighted avg       0.70      0.70      0.70      1000\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "yEstN = model0Norway.predict(xTestN)\n",
    "plotConfusionMatrix(yTestN, yEstN, norwayColors, title=\"Norway flag confusion matrix\")\n",
    "print(metrics.classification_report(yTestN, yEstN))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Norway flag problem is a little more challenging as the there are more borders between the three colors.\n",
    "\n",
    "Let's investigate the best parameters using accuracy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hyper-parameter search using grid on Norway flag\n",
    "\n",
    "A grid search is performed on the two tree partitioner stop conditions : on minimum impurity, and on minimum leave size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "leafSizeRange = range(4, 21)\n",
    "entropyRange = np.logspace(-3, 0, 10)\n",
    "accuracy = np.zeros((len(leafSizeRange), len(entropyRange)))\n",
    "for i,leafMinSize in enumerate(leafSizeRange):\n",
    "    for j,entropy in enumerate(entropyRange): \n",
    "        model1Norway = MyTreeClassifier(leafMinSize=leafMinSize, entropyMin=entropy)\n",
    "        model1Norway.fit(xTrainN, yTrainN, range(3))\n",
    "        yEstN1 = model1Norway.predict(xTestN)\n",
    "        accuracy[i,j] = np.mean(yEstN1 == yTestN)\n",
    "\n",
    "bestParams = np.argmax(accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1, figsize=(10, 5))\n",
    "for j,h in enumerate(entropyRange):\n",
    "    ax.plot(leafSizeRange, accuracy[:,j], label = \"Min entropy %.2e\" % h)\n",
    "ax.legend()\n",
    "ax.set_title(\"Accuracy as function of min leaf size and min entropy\")\n",
    "ax.set_xlabel(\"Leaf minimum size\")\n",
    "ax.set_ylabel(\"Accuracy\")\n",
    "ax.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Norway flag classification with leaf min size = 13, min entropy = 0.215\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.84      0.81      0.82       319\n",
      "         1.0       0.60      0.67      0.63       366\n",
      "         2.0       0.73      0.65      0.69       315\n",
      "\n",
      "    accuracy                           0.71      1000\n",
      "   macro avg       0.72      0.71      0.71      1000\n",
      "weighted avg       0.71      0.71      0.71      1000\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x288 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "leafMinSize = leafSizeRange[bestParams // len(leafSizeRange)]\n",
    "entropyMin = entropyRange[bestParams % len(leafSizeRange)]\n",
    "model1Norway = MyTreeClassifier(leafMinSize=leafMinSize, entropyMin=entropyMin)\n",
    "processedListN1 = model1Norway.fit(xTrainN, yTrainN, range(3))\n",
    "yEstN1 = model1Norway.predict(xTestN)\n",
    "\n",
    "print(\"Norway flag classification with leaf min size = %d, min entropy = %.3f\" % (leafMinSize, entropyMin))\n",
    "print(metrics.classification_report(yTestN, yEstN1))\n",
    "\n",
    "fig, axes = plt.subplots(1, 3, figsize=(16, 4))\n",
    "plotConfusionMatrix(yTestN, yEstN1, norwayColors, ax=axes[0], title=\"Norway flag confusion matrix\")\n",
    "axes[1].set_title(\"Partitions of the decision tree\")\n",
    "axes[1].set_facecolor((0.8, 0.8, 0.8))\n",
    "for targetDepth in range(1, 10):\n",
    "    for n in filter(lambda n: n.depth == targetDepth, processedListN1):\n",
    "        plotNodeArea(n, norwayColors, axes[1])\n",
    "flagPlot(xTestN, yTestN, \"Norway flag with true test samples\", norwayColors, ax=axes[2])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The minimum entropy stop criteria has the advantage of avoid very fine and small partitions and thus decrease the need of consolidating the leafs (also called pruning). However it may also prevent useful splits at the early stages as seen on the large white rectangles at right hand of above partition map.\n",
    "\n",
    "Comparing the partition map based on the training samples and the plot of the test samples, we see that there is some overfitting."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Classification tree with Scikit-Learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeClassifier(class_weight=None, criterion='entropy', max_depth=None,\n",
       "                       max_features=None, max_leaf_nodes=None,\n",
       "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                       min_samples_leaf=8, min_samples_split=2,\n",
       "                       min_weight_fraction_leaf=0.0, presort=False,\n",
       "                       random_state=None, splitter='best')"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model2Norway = tree.DecisionTreeClassifier(min_samples_leaf=8, criterion='entropy')\n",
    "model2Norway.fit(xTrainN, yTrainN)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Norway flag classification with leaf min size = 8\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.87      0.82      0.84       319\n",
      "         1.0       0.62      0.60      0.61       366\n",
      "         2.0       0.67      0.74      0.71       315\n",
      "\n",
      "    accuracy                           0.71      1000\n",
      "   macro avg       0.72      0.72      0.72      1000\n",
      "weighted avg       0.72      0.71      0.71      1000\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "yEstN2 = model2Norway.predict(xTestN)\n",
    "print(\"Norway flag classification with leaf min size = 8\")\n",
    "print(metrics.classification_report(yTestN, yEstN2))\n",
    "\n",
    "plotConfusionMatrix(yTestN, yEstN2, norwayColors)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualization of the decision tree\n",
    "\n",
    "Scikit-Learn provides a graph view of the Tree... not that easy to visualize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<svg height=\"1385pt\" viewBox=\"0.00 0.00 4489.00 1385.00\" width=\"4489pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g class=\"graph\" id=\"graph0\" transform=\"scale(1 1) rotate(0) translate(4 1381)\">\n",
       "<title>Tree</title>\n",
       "<polygon fill=\"#ffffff\" points=\"-4,4 -4,-1381 4485,-1381 4485,4 -4,4\" stroke=\"transparent\"/>\n",
       "<!-- 0 -->\n",
       "<g class=\"node\" id=\"node1\">\n",
       "<title>0</title>\n",
       "<polygon fill=\"#fbfffd\" points=\"1506.5,-1377 1358.5,-1377 1358.5,-1294 1506.5,-1294 1506.5,-1377\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1396.5\" y=\"-1361.8\">x_1 ≤ -0.418</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1387.5\" y=\"-1346.8\">entropy = 1.583</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1388\" y=\"-1331.8\">samples = 1000</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1366.5\" y=\"-1316.8\">value = [310, 351, 339]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1395.5\" y=\"-1301.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 1 -->\n",
       "<g class=\"node\" id=\"node2\">\n",
       "<title>1</title>\n",
       "<polygon fill=\"#f4c9aa\" points=\"1050.5,-1258 916.5,-1258 916.5,-1175 1050.5,-1175 1050.5,-1258\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"947.5\" y=\"-1242.8\">x_2 ≤ -0.375</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"938.5\" y=\"-1227.8\">entropy = 1.401</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"942.5\" y=\"-1212.8\">samples = 281</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"924.5\" y=\"-1197.8\">value = [161, 71, 49]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"953.5\" y=\"-1182.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 0&#45;&gt;1 -->\n",
       "<g class=\"edge\" id=\"edge1\">\n",
       "<title>0-&gt;1</title>\n",
       "<path d=\"M1358.3568,-1315.8496C1276.0566,-1294.0373 1143.8633,-1259.0016 1060.4493,-1236.8941\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1061.3397,-1233.5093 1050.7767,-1234.3306 1059.5463,-1240.2757 1061.3397,-1233.5093\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1063.3356\" y=\"-1252.2471\">True</text>\n",
       "</g>\n",
       "<!-- 38 -->\n",
       "<g class=\"node\" id=\"node39\">\n",
       "<title>38</title>\n",
       "<polygon fill=\"#fcfafe\" points=\"2024.5,-1258 1876.5,-1258 1876.5,-1175 2024.5,-1175 2024.5,-1258\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1916.5\" y=\"-1242.8\">x_1 ≤ 0.372</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1905.5\" y=\"-1227.8\">entropy = 1.529</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1909.5\" y=\"-1212.8\">samples = 719</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1884.5\" y=\"-1197.8\">value = [149, 280, 290]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1915.5\" y=\"-1182.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 0&#45;&gt;38 -->\n",
       "<g class=\"edge\" id=\"edge38\">\n",
       "<title>0-&gt;38</title>\n",
       "<path d=\"M1506.7168,-1318.4502C1601.9189,-1296.5794 1766.6235,-1258.7419 1866.3784,-1235.8252\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1867.4584,-1239.1684 1876.4209,-1233.5182 1865.8911,-1232.3461 1867.4584,-1239.1684\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1863.15\" y=\"-1251.005\">False</text>\n",
       "</g>\n",
       "<!-- 2 -->\n",
       "<g class=\"node\" id=\"node3\">\n",
       "<title>2</title>\n",
       "<polygon fill=\"#e88f4f\" points=\"503,-1139 390,-1139 390,-1056 503,-1056 503,-1139\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"410.5\" y=\"-1123.8\">x_1 ≤ -0.582</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"401.5\" y=\"-1108.8\">entropy = 0.473</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"409\" y=\"-1093.8\">samples = 89</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"398\" y=\"-1078.8\">value = [80, 9, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"416.5\" y=\"-1063.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 1&#45;&gt;2 -->\n",
       "<g class=\"edge\" id=\"edge2\">\n",
       "<title>1-&gt;2</title>\n",
       "<path d=\"M916.3036,-1201.6092C812.9443,-1178.7046 616.1299,-1135.0902 513.1196,-1112.263\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"513.6683,-1108.7998 503.1479,-1110.0533 512.1538,-1115.634 513.6683,-1108.7998\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 11 -->\n",
       "<g class=\"node\" id=\"node12\">\n",
       "<title>11</title>\n",
       "<polygon fill=\"#fbede2\" points=\"1047,-1139 920,-1139 920,-1056 1047,-1056 1047,-1139\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"949.5\" y=\"-1123.8\">x_2 ≤ 0.341</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"938.5\" y=\"-1108.8\">entropy = 1.555</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"942.5\" y=\"-1093.8\">samples = 192</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"928\" y=\"-1078.8\">value = [81, 62, 49]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"953.5\" y=\"-1063.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 1&#45;&gt;11 -->\n",
       "<g class=\"edge\" id=\"edge11\">\n",
       "<title>1-&gt;11</title>\n",
       "<path d=\"M983.5,-1174.8796C983.5,-1166.6838 983.5,-1157.9891 983.5,-1149.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"987.0001,-1149.298 983.5,-1139.2981 980.0001,-1149.2981 987.0001,-1149.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 3 -->\n",
       "<g class=\"node\" id=\"node4\">\n",
       "<title>3</title>\n",
       "<polygon fill=\"#e6853f\" points=\"308,-1020 195,-1020 195,-937 308,-937 308,-1020\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"215.5\" y=\"-1004.8\">x_1 ≤ -0.884</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"206.5\" y=\"-989.8\">entropy = 0.201</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"214\" y=\"-974.8\">samples = 64</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"203\" y=\"-959.8\">value = [62, 2, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"221.5\" y=\"-944.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 2&#45;&gt;3 -->\n",
       "<g class=\"edge\" id=\"edge3\">\n",
       "<title>2-&gt;3</title>\n",
       "<path d=\"M389.836,-1062.9204C367.0017,-1048.9856 340.52,-1032.825 316.8759,-1018.3961\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"318.6647,-1015.3874 308.3054,-1013.1658 315.0182,-1021.3627 318.6647,-1015.3874\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 8 -->\n",
       "<g class=\"node\" id=\"node9\">\n",
       "<title>8</title>\n",
       "<polygon fill=\"#efb286\" points=\"503,-1020 390,-1020 390,-937 503,-937 503,-1020\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"410.5\" y=\"-1004.8\">x_1 ≤ -0.497</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"401.5\" y=\"-989.8\">entropy = 0.855</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"409\" y=\"-974.8\">samples = 25</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"398\" y=\"-959.8\">value = [18, 7, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"416.5\" y=\"-944.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 2&#45;&gt;8 -->\n",
       "<g class=\"edge\" id=\"edge8\">\n",
       "<title>2-&gt;8</title>\n",
       "<path d=\"M446.5,-1055.8796C446.5,-1047.6838 446.5,-1038.9891 446.5,-1030.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"450.0001,-1030.298 446.5,-1020.2981 443.0001,-1030.2981 450.0001,-1030.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 4 -->\n",
       "<g class=\"node\" id=\"node5\">\n",
       "<title>4</title>\n",
       "<polygon fill=\"#e88f4f\" points=\"177,-901 64,-901 64,-818 177,-818 177,-901\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"88\" y=\"-885.8\">x_1 ≤ -0.94</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"75.5\" y=\"-870.8\">entropy = 0.469</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"83\" y=\"-855.8\">samples = 20</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"72\" y=\"-840.8\">value = [18, 2, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"90.5\" y=\"-825.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 3&#45;&gt;4 -->\n",
       "<g class=\"edge\" id=\"edge4\">\n",
       "<title>3-&gt;4</title>\n",
       "<path d=\"M205.6826,-936.8796C195.4706,-927.6031 184.555,-917.6874 174.0624,-908.1559\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"176.2684,-905.4313 166.513,-901.2981 171.5616,-910.6127 176.2684,-905.4313\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 7 -->\n",
       "<g class=\"node\" id=\"node8\">\n",
       "<title>7</title>\n",
       "<polygon fill=\"#e58139\" points=\"308,-893.5 195,-893.5 195,-825.5 308,-825.5 308,-893.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"213.5\" y=\"-878.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"214\" y=\"-863.3\">samples = 44</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"203\" y=\"-848.3\">value = [44, 0, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"221.5\" y=\"-833.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 3&#45;&gt;7 -->\n",
       "<g class=\"edge\" id=\"edge7\">\n",
       "<title>3-&gt;7</title>\n",
       "<path d=\"M251.5,-936.8796C251.5,-926.2134 251.5,-914.7021 251.5,-903.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"255.0001,-903.8149 251.5,-893.8149 248.0001,-903.815 255.0001,-903.8149\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 5 -->\n",
       "<g class=\"node\" id=\"node6\">\n",
       "<title>5</title>\n",
       "<polygon fill=\"#e58139\" points=\"113,-774.5 0,-774.5 0,-706.5 113,-706.5 113,-774.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"18.5\" y=\"-759.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"19\" y=\"-744.3\">samples = 12</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"8\" y=\"-729.3\">value = [12, 0, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"26.5\" y=\"-714.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 4&#45;&gt;5 -->\n",
       "<g class=\"edge\" id=\"edge5\">\n",
       "<title>4-&gt;5</title>\n",
       "<path d=\"M98.1159,-817.8796C92.202,-806.8835 85.8051,-794.9893 79.842,-783.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"82.7742,-781.9642 74.9551,-774.8149 76.6092,-785.2799 82.7742,-781.9642\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 6 -->\n",
       "<g class=\"node\" id=\"node7\">\n",
       "<title>6</title>\n",
       "<polygon fill=\"#eeab7b\" points=\"237.5,-774.5 131.5,-774.5 131.5,-706.5 237.5,-706.5 237.5,-774.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"139.5\" y=\"-759.3\">entropy = 0.811</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"150.5\" y=\"-744.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"139.5\" y=\"-729.3\">value = [6, 2, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"154.5\" y=\"-714.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 4&#45;&gt;6 -->\n",
       "<g class=\"edge\" id=\"edge6\">\n",
       "<title>4-&gt;6</title>\n",
       "<path d=\"M142.8841,-817.8796C148.798,-806.8835 155.1949,-794.9893 161.158,-783.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"164.3908,-785.2799 166.0449,-774.8149 158.2258,-781.9642 164.3908,-785.2799\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 9 -->\n",
       "<g class=\"node\" id=\"node10\">\n",
       "<title>9</title>\n",
       "<polygon fill=\"#eca36f\" points=\"439,-893.5 326,-893.5 326,-825.5 439,-825.5 439,-893.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"341\" y=\"-878.3\">entropy = 0.75</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"345\" y=\"-863.3\">samples = 14</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"334\" y=\"-848.3\">value = [11, 3, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"352.5\" y=\"-833.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 8&#45;&gt;9 -->\n",
       "<g class=\"edge\" id=\"edge9\">\n",
       "<title>8-&gt;9</title>\n",
       "<path d=\"M424.1159,-936.8796C418.202,-925.8835 411.8051,-913.9893 405.842,-902.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"408.7742,-900.9642 400.9551,-893.8149 402.6092,-904.2799 408.7742,-900.9642\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 10 -->\n",
       "<g class=\"node\" id=\"node11\">\n",
       "<title>10</title>\n",
       "<polygon fill=\"#f4c9aa\" points=\"563.5,-893.5 457.5,-893.5 457.5,-825.5 563.5,-825.5 563.5,-893.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"465.5\" y=\"-878.3\">entropy = 0.946</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"473\" y=\"-863.3\">samples = 11</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"465.5\" y=\"-848.3\">value = [7, 4, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"480.5\" y=\"-833.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 8&#45;&gt;10 -->\n",
       "<g class=\"edge\" id=\"edge10\">\n",
       "<title>8-&gt;10</title>\n",
       "<path d=\"M468.8841,-936.8796C474.798,-925.8835 481.1949,-913.9893 487.158,-902.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"490.3908,-904.2799 492.0449,-893.8149 484.2258,-900.9642 490.3908,-904.2799\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 12 -->\n",
       "<g class=\"node\" id=\"node13\">\n",
       "<title>12</title>\n",
       "<polygon fill=\"#f8f4fe\" points=\"915.5,-1020 795.5,-1020 795.5,-937 915.5,-937 915.5,-1020\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"819.5\" y=\"-1004.8\">x_2 ≤ -0.166</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"810.5\" y=\"-989.8\">entropy = 1.236</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"814.5\" y=\"-974.8\">samples = 100</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"803.5\" y=\"-959.8\">value = [5, 46, 49]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"820.5\" y=\"-944.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 11&#45;&gt;12 -->\n",
       "<g class=\"edge\" id=\"edge12\">\n",
       "<title>11-&gt;12</title>\n",
       "<path d=\"M938.7318,-1055.8796C928.7537,-1046.6031 918.0881,-1036.6874 907.8358,-1027.1559\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"910.1663,-1024.5436 900.4593,-1020.2981 905.4,-1029.6704 910.1663,-1024.5436\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 29 -->\n",
       "<g class=\"node\" id=\"node30\">\n",
       "<title>29</title>\n",
       "<polygon fill=\"#ea9c63\" points=\"1212.5,-1020 1092.5,-1020 1092.5,-937 1212.5,-937 1212.5,-1020\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1116.5\" y=\"-1004.8\">x_1 ≤ -0.535</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1107.5\" y=\"-989.8\">entropy = 0.667</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1115\" y=\"-974.8\">samples = 92</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1100.5\" y=\"-959.8\">value = [76, 16, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1122.5\" y=\"-944.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 11&#45;&gt;29 -->\n",
       "<g class=\"edge\" id=\"edge29\">\n",
       "<title>11-&gt;29</title>\n",
       "<path d=\"M1042.608,-1055.8796C1056.3734,-1046.1868 1071.1299,-1035.7961 1085.2193,-1025.8752\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1087.2641,-1028.716 1093.4254,-1020.0969 1083.2339,-1022.9926 1087.2641,-1028.716\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 13 -->\n",
       "<g class=\"node\" id=\"node14\">\n",
       "<title>13</title>\n",
       "<polygon fill=\"#5dea98\" points=\"695,-901 582,-901 582,-818 695,-818 695,-901\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"602.5\" y=\"-885.8\">x_1 ≤ -0.818</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"593.5\" y=\"-870.8\">entropy = 0.817</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"601\" y=\"-855.8\">samples = 24</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"590\" y=\"-840.8\">value = [2, 20, 2]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"601.5\" y=\"-825.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 12&#45;&gt;13 -->\n",
       "<g class=\"edge\" id=\"edge13\">\n",
       "<title>12-&gt;13</title>\n",
       "<path d=\"M795.3416,-945.5099C766.9239,-929.926 732.9449,-911.2924 703.8909,-895.3595\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"705.4949,-892.2475 695.0439,-890.5079 702.1291,-898.3851 705.4949,-892.2475\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 18 -->\n",
       "<g class=\"node\" id=\"node19\">\n",
       "<title>18</title>\n",
       "<polygon fill=\"#caacf4\" points=\"915.5,-901 795.5,-901 795.5,-818 915.5,-818 915.5,-901\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"821.5\" y=\"-885.8\">x_2 ≤ 0.134</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"810.5\" y=\"-870.8\">entropy = 1.142</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"818\" y=\"-855.8\">samples = 76</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"803.5\" y=\"-840.8\">value = [3, 26, 47]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"820.5\" y=\"-825.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 12&#45;&gt;18 -->\n",
       "<g class=\"edge\" id=\"edge18\">\n",
       "<title>12-&gt;18</title>\n",
       "<path d=\"M855.5,-936.8796C855.5,-928.6838 855.5,-919.9891 855.5,-911.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"859.0001,-911.298 855.5,-901.2981 852.0001,-911.2981 859.0001,-911.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 14 -->\n",
       "<g class=\"node\" id=\"node15\">\n",
       "<title>14</title>\n",
       "<polygon fill=\"#7beeab\" points=\"522.5,-774.5 416.5,-774.5 416.5,-706.5 522.5,-706.5 522.5,-774.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"424.5\" y=\"-759.3\">entropy = 0.811</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"435.5\" y=\"-744.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"424.5\" y=\"-729.3\">value = [2, 6, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"432.5\" y=\"-714.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 13&#45;&gt;14 -->\n",
       "<g class=\"edge\" id=\"edge14\">\n",
       "<title>13-&gt;14</title>\n",
       "<path d=\"M581.7553,-819.5437C563.9555,-807.0101 544.3073,-793.1749 526.4905,-780.6294\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"528.0989,-777.4813 517.9074,-774.5857 524.0687,-783.2048 528.0989,-777.4813\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 15 -->\n",
       "<g class=\"node\" id=\"node16\">\n",
       "<title>15</title>\n",
       "<polygon fill=\"#55e993\" points=\"654,-782 541,-782 541,-699 654,-699 654,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"565\" y=\"-766.8\">x_2 ≤ -0.25</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"552.5\" y=\"-751.8\">entropy = 0.544</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"560\" y=\"-736.8\">samples = 16</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"549\" y=\"-721.8\">value = [0, 14, 2]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"560.5\" y=\"-706.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 13&#45;&gt;15 -->\n",
       "<g class=\"edge\" id=\"edge15\">\n",
       "<title>13-&gt;15</title>\n",
       "<path d=\"M624.1602,-817.8796C621.2744,-809.5037 618.209,-800.6067 615.2237,-791.942\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"618.4676,-790.6125 611.901,-782.2981 611.8494,-792.8928 618.4676,-790.6125\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 16 -->\n",
       "<g class=\"node\" id=\"node17\">\n",
       "<title>16</title>\n",
       "<polygon fill=\"#55e993\" points=\"435.5,-655.5 329.5,-655.5 329.5,-587.5 435.5,-587.5 435.5,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"337.5\" y=\"-640.3\">entropy = 0.544</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"348.5\" y=\"-625.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"337.5\" y=\"-610.3\">value = [0, 7, 1]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"345.5\" y=\"-595.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 15&#45;&gt;16 -->\n",
       "<g class=\"edge\" id=\"edge16\">\n",
       "<title>15-&gt;16</title>\n",
       "<path d=\"M540.9754,-704.1085C537.7989,-702.3303 534.6266,-700.6152 531.5,-699 494.3215,-679.794 481.4006,-682.7346 444.5,-663 443.1793,-662.2937 441.8515,-661.5666 440.5198,-660.8219\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"442.1128,-657.6993 431.7107,-655.6878 438.5881,-663.7471 442.1128,-657.6993\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 17 -->\n",
       "<g class=\"node\" id=\"node18\">\n",
       "<title>17</title>\n",
       "<polygon fill=\"#55e993\" points=\"559.5,-655.5 453.5,-655.5 453.5,-587.5 559.5,-587.5 559.5,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"461.5\" y=\"-640.3\">entropy = 0.544</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"472.5\" y=\"-625.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"461.5\" y=\"-610.3\">value = [0, 7, 1]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"469.5\" y=\"-595.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 15&#45;&gt;17 -->\n",
       "<g class=\"edge\" id=\"edge17\">\n",
       "<title>15-&gt;17</title>\n",
       "<path d=\"M565.6726,-698.8796C557.0116,-687.5536 547.6219,-675.2748 538.928,-663.9058\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"541.5956,-661.6324 532.7408,-655.8149 536.0351,-665.8846 541.5956,-661.6324\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 19 -->\n",
       "<g class=\"node\" id=\"node20\">\n",
       "<title>19</title>\n",
       "<polygon fill=\"#8d4ce8\" points=\"822,-782 709,-782 709,-699 822,-699 822,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"729.5\" y=\"-766.8\">x_1 ≤ -0.766</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"720.5\" y=\"-751.8\">entropy = 0.433</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"728\" y=\"-736.8\">samples = 45</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"717\" y=\"-721.8\">value = [0, 4, 41]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"730.5\" y=\"-706.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 18&#45;&gt;19 -->\n",
       "<g class=\"edge\" id=\"edge19\">\n",
       "<title>18-&gt;19</title>\n",
       "<path d=\"M824.0224,-817.8796C817.3471,-809.0534 810.2342,-799.6485 803.3504,-790.5466\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"805.9357,-788.1626 797.112,-782.2981 800.3526,-792.3852 805.9357,-788.1626\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 24 -->\n",
       "<g class=\"node\" id=\"node25\">\n",
       "<title>24</title>\n",
       "<polygon fill=\"#80eeae\" points=\"953,-782 840,-782 840,-699 953,-699 953,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"860.5\" y=\"-766.8\">x_1 ≤ -0.774</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"851.5\" y=\"-751.8\">entropy = 1.136</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"859\" y=\"-736.8\">samples = 31</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"848\" y=\"-721.8\">value = [3, 22, 6]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"859.5\" y=\"-706.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 18&#45;&gt;24 -->\n",
       "<g class=\"edge\" id=\"edge24\">\n",
       "<title>18-&gt;24</title>\n",
       "<path d=\"M869.8398,-817.8796C872.7256,-809.5037 875.791,-800.6067 878.7763,-791.942\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"882.1506,-792.8928 882.099,-782.2981 875.5324,-790.6125 882.1506,-792.8928\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 20 -->\n",
       "<g class=\"node\" id=\"node21\">\n",
       "<title>20</title>\n",
       "<polygon fill=\"#8139e5\" points=\"691,-655.5 578,-655.5 578,-587.5 691,-587.5 691,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"596.5\" y=\"-640.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"597\" y=\"-625.3\">samples = 19</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"586\" y=\"-610.3\">value = [0, 0, 19]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"599.5\" y=\"-595.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 19&#45;&gt;20 -->\n",
       "<g class=\"edge\" id=\"edge20\">\n",
       "<title>19-&gt;20</title>\n",
       "<path d=\"M719.6826,-698.8796C706.7303,-687.1138 692.646,-674.3197 679.7286,-662.5855\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"682.0306,-659.9482 672.2753,-655.8149 677.3239,-665.1296 682.0306,-659.9482\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 21 -->\n",
       "<g class=\"node\" id=\"node22\">\n",
       "<title>21</title>\n",
       "<polygon fill=\"#985dea\" points=\"822,-663 709,-663 709,-580 822,-580 822,-663\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"729.5\" y=\"-647.8\">x_1 ≤ -0.596</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"720.5\" y=\"-632.8\">entropy = 0.619</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"728\" y=\"-617.8\">samples = 26</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"717\" y=\"-602.8\">value = [0, 4, 22]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"730.5\" y=\"-587.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 19&#45;&gt;21 -->\n",
       "<g class=\"edge\" id=\"edge21\">\n",
       "<title>19-&gt;21</title>\n",
       "<path d=\"M765.5,-698.8796C765.5,-690.6838 765.5,-681.9891 765.5,-673.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"769.0001,-673.298 765.5,-663.2981 762.0001,-673.2981 769.0001,-673.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 22 -->\n",
       "<g class=\"node\" id=\"node23\">\n",
       "<title>22</title>\n",
       "<polygon fill=\"#af81ee\" points=\"695,-536.5 582,-536.5 582,-468.5 695,-468.5 695,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"593.5\" y=\"-521.3\">entropy = 0.837</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"601\" y=\"-506.3\">samples = 15</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"590\" y=\"-491.3\">value = [0, 4, 11]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"603.5\" y=\"-476.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 21&#45;&gt;22 -->\n",
       "<g class=\"edge\" id=\"edge22\">\n",
       "<title>21-&gt;22</title>\n",
       "<path d=\"M721.0816,-579.8796C708.6421,-568.2237 695.1257,-555.5587 682.699,-543.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"684.8121,-541.0985 675.1218,-536.8149 680.0259,-546.2065 684.8121,-541.0985\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 23 -->\n",
       "<g class=\"node\" id=\"node24\">\n",
       "<title>23</title>\n",
       "<polygon fill=\"#8139e5\" points=\"826,-536.5 713,-536.5 713,-468.5 826,-468.5 826,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"731.5\" y=\"-521.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"732\" y=\"-506.3\">samples = 11</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"721\" y=\"-491.3\">value = [0, 0, 11]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"734.5\" y=\"-476.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 21&#45;&gt;23 -->\n",
       "<g class=\"edge\" id=\"edge23\">\n",
       "<title>21-&gt;23</title>\n",
       "<path d=\"M766.899,-579.8796C767.2575,-569.2134 767.6445,-557.7021 768.0075,-546.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"771.5085,-546.9269 768.3466,-536.8149 764.5125,-546.6917 771.5085,-546.9269\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 25 -->\n",
       "<g class=\"node\" id=\"node26\">\n",
       "<title>25</title>\n",
       "<polygon fill=\"#47e78a\" points=\"953,-655.5 840,-655.5 840,-587.5 953,-587.5 953,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"851.5\" y=\"-640.3\">entropy = 0.353</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"859\" y=\"-625.3\">samples = 15</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"848\" y=\"-610.3\">value = [1, 14, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"859.5\" y=\"-595.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 24&#45;&gt;25 -->\n",
       "<g class=\"edge\" id=\"edge25\">\n",
       "<title>24-&gt;25</title>\n",
       "<path d=\"M896.5,-698.8796C896.5,-688.2134 896.5,-676.7021 896.5,-665.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"900.0001,-665.8149 896.5,-655.8149 893.0001,-665.815 900.0001,-665.8149\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 26 -->\n",
       "<g class=\"node\" id=\"node27\">\n",
       "<title>26</title>\n",
       "<polygon fill=\"#d7fae6\" points=\"1077.5,-663 971.5,-663 971.5,-580 1077.5,-580 1077.5,-663\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"990.5\" y=\"-647.8\">x_2 ≤ 0.244</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"979.5\" y=\"-632.8\">entropy = 1.406</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"987\" y=\"-617.8\">samples = 16</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"979.5\" y=\"-602.8\">value = [2, 8, 6]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"987.5\" y=\"-587.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 24&#45;&gt;26 -->\n",
       "<g class=\"edge\" id=\"edge26\">\n",
       "<title>24-&gt;26</title>\n",
       "<path d=\"M941.2682,-698.8796C951.2463,-689.6031 961.9119,-679.6874 972.1642,-670.1559\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"974.6,-672.6704 979.5407,-663.2981 969.8337,-667.5436 974.6,-672.6704\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 27 -->\n",
       "<g class=\"node\" id=\"node28\">\n",
       "<title>27</title>\n",
       "<polygon fill=\"#9cf2c0\" points=\"953.5,-536.5 847.5,-536.5 847.5,-468.5 953.5,-468.5 953.5,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"855.5\" y=\"-521.3\">entropy = 1.299</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"866.5\" y=\"-506.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"855.5\" y=\"-491.3\">value = [1, 5, 2]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"863.5\" y=\"-476.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 26&#45;&gt;27 -->\n",
       "<g class=\"edge\" id=\"edge27\">\n",
       "<title>26-&gt;27</title>\n",
       "<path d=\"M981.1308,-579.8796C968.9852,-568.2237 955.7881,-555.5587 943.6549,-543.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"945.8953,-541.2138 936.2567,-536.8149 941.0484,-546.2643 945.8953,-541.2138\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 28 -->\n",
       "<g class=\"node\" id=\"node29\">\n",
       "<title>28</title>\n",
       "<polygon fill=\"#e6d7fa\" points=\"1077.5,-536.5 971.5,-536.5 971.5,-468.5 1077.5,-468.5 1077.5,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"979.5\" y=\"-521.3\">entropy = 1.406</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"990.5\" y=\"-506.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"979.5\" y=\"-491.3\">value = [1, 3, 4]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"989.5\" y=\"-476.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 26&#45;&gt;28 -->\n",
       "<g class=\"edge\" id=\"edge28\">\n",
       "<title>26-&gt;28</title>\n",
       "<path d=\"M1024.5,-579.8796C1024.5,-569.2134 1024.5,-557.7021 1024.5,-546.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1028.0001,-546.8149 1024.5,-536.8149 1021.0001,-546.815 1028.0001,-546.8149\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 30 -->\n",
       "<g class=\"node\" id=\"node31\">\n",
       "<title>30</title>\n",
       "<polygon fill=\"#e89253\" points=\"1209,-901 1096,-901 1096,-818 1209,-818 1209,-901\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1118.5\" y=\"-885.8\">x_2 ≤ 0.561</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1111\" y=\"-870.8\">entropy = 0.52</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1115\" y=\"-855.8\">samples = 77</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1104\" y=\"-840.8\">value = [68, 9, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1122.5\" y=\"-825.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 29&#45;&gt;30 -->\n",
       "<g class=\"edge\" id=\"edge30\">\n",
       "<title>29-&gt;30</title>\n",
       "<path d=\"M1152.5,-936.8796C1152.5,-928.6838 1152.5,-919.9891 1152.5,-911.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1156.0001,-911.298 1152.5,-901.2981 1149.0001,-911.2981 1156.0001,-911.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 37 -->\n",
       "<g class=\"node\" id=\"node38\">\n",
       "<title>37</title>\n",
       "<polygon fill=\"#fcefe6\" points=\"1333.5,-893.5 1227.5,-893.5 1227.5,-825.5 1333.5,-825.5 1333.5,-893.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1235.5\" y=\"-878.3\">entropy = 0.997</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1243\" y=\"-863.3\">samples = 15</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1235.5\" y=\"-848.3\">value = [8, 7, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1250.5\" y=\"-833.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 29&#45;&gt;37 -->\n",
       "<g class=\"edge\" id=\"edge37\">\n",
       "<title>29-&gt;37</title>\n",
       "<path d=\"M1197.2682,-936.8796C1209.8056,-925.2237 1223.4284,-912.5587 1235.953,-900.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1238.6491,-903.1872 1243.5898,-893.8149 1233.8828,-898.0605 1238.6491,-903.1872\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 31 -->\n",
       "<g class=\"node\" id=\"node32\">\n",
       "<title>31</title>\n",
       "<polygon fill=\"#f1ba92\" points=\"1209,-782 1096,-782 1096,-699 1209,-699 1209,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1116.5\" y=\"-766.8\">x_1 ≤ -0.644</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1107.5\" y=\"-751.8\">entropy = 0.894</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1115\" y=\"-736.8\">samples = 29</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1104\" y=\"-721.8\">value = [20, 9, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1122.5\" y=\"-706.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 30&#45;&gt;31 -->\n",
       "<g class=\"edge\" id=\"edge31\">\n",
       "<title>30-&gt;31</title>\n",
       "<path d=\"M1152.5,-817.8796C1152.5,-809.6838 1152.5,-800.9891 1152.5,-792.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1156.0001,-792.298 1152.5,-782.2981 1149.0001,-792.2981 1156.0001,-792.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 36 -->\n",
       "<g class=\"node\" id=\"node37\">\n",
       "<title>36</title>\n",
       "<polygon fill=\"#e58139\" points=\"1340,-774.5 1227,-774.5 1227,-706.5 1340,-706.5 1340,-774.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1245.5\" y=\"-759.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1246\" y=\"-744.3\">samples = 48</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1235\" y=\"-729.3\">value = [48, 0, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1253.5\" y=\"-714.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 30&#45;&gt;36 -->\n",
       "<g class=\"edge\" id=\"edge36\">\n",
       "<title>30-&gt;36</title>\n",
       "<path d=\"M1198.3174,-817.8796C1211.2697,-806.1138 1225.354,-793.3197 1238.2714,-781.5855\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1240.6761,-784.1296 1245.7247,-774.8149 1235.9694,-778.9482 1240.6761,-784.1296\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 32 -->\n",
       "<g class=\"node\" id=\"node33\">\n",
       "<title>32</title>\n",
       "<polygon fill=\"#f5cfb3\" points=\"1209,-663 1096,-663 1096,-580 1209,-580 1209,-663\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1116.5\" y=\"-647.8\">x_1 ≤ -0.782</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1107.5\" y=\"-632.8\">entropy = 0.959</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1115\" y=\"-617.8\">samples = 21</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1104\" y=\"-602.8\">value = [13, 8, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1122.5\" y=\"-587.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 31&#45;&gt;32 -->\n",
       "<g class=\"edge\" id=\"edge32\">\n",
       "<title>31-&gt;32</title>\n",
       "<path d=\"M1152.5,-698.8796C1152.5,-690.6838 1152.5,-681.9891 1152.5,-673.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1156.0001,-673.298 1152.5,-663.2981 1149.0001,-673.2981 1156.0001,-673.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 35 -->\n",
       "<g class=\"node\" id=\"node36\">\n",
       "<title>35</title>\n",
       "<polygon fill=\"#e99355\" points=\"1333.5,-655.5 1227.5,-655.5 1227.5,-587.5 1333.5,-587.5 1333.5,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1235.5\" y=\"-640.3\">entropy = 0.544</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1246.5\" y=\"-625.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1235.5\" y=\"-610.3\">value = [7, 1, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1250.5\" y=\"-595.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 31&#45;&gt;35 -->\n",
       "<g class=\"edge\" id=\"edge35\">\n",
       "<title>31-&gt;35</title>\n",
       "<path d=\"M1197.2682,-698.8796C1209.8056,-687.2237 1223.4284,-674.5587 1235.953,-662.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1238.6491,-665.1872 1243.5898,-655.8149 1233.8828,-660.0605 1238.6491,-665.1872\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 33 -->\n",
       "<g class=\"node\" id=\"node34\">\n",
       "<title>33</title>\n",
       "<polygon fill=\"#eda774\" points=\"1209,-536.5 1096,-536.5 1096,-468.5 1209,-468.5 1209,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1107.5\" y=\"-521.3\">entropy = 0.779</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1115\" y=\"-506.3\">samples = 13</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1104\" y=\"-491.3\">value = [10, 3, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1122.5\" y=\"-476.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 32&#45;&gt;33 -->\n",
       "<g class=\"edge\" id=\"edge33\">\n",
       "<title>32-&gt;33</title>\n",
       "<path d=\"M1152.5,-579.8796C1152.5,-569.2134 1152.5,-557.7021 1152.5,-546.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1156.0001,-546.8149 1152.5,-536.8149 1149.0001,-546.815 1156.0001,-546.8149\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 34 -->\n",
       "<g class=\"node\" id=\"node35\">\n",
       "<title>34</title>\n",
       "<polygon fill=\"#b0f5cd\" points=\"1333.5,-536.5 1227.5,-536.5 1227.5,-468.5 1333.5,-468.5 1333.5,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1235.5\" y=\"-521.3\">entropy = 0.954</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1246.5\" y=\"-506.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1235.5\" y=\"-491.3\">value = [3, 5, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1243.5\" y=\"-476.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 32&#45;&gt;34 -->\n",
       "<g class=\"edge\" id=\"edge34\">\n",
       "<title>32-&gt;34</title>\n",
       "<path d=\"M1197.2682,-579.8796C1209.8056,-568.2237 1223.4284,-555.5587 1235.953,-543.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1238.6491,-546.1872 1243.5898,-536.8149 1233.8828,-541.0605 1238.6491,-546.1872\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 39 -->\n",
       "<g class=\"node\" id=\"node40\">\n",
       "<title>39</title>\n",
       "<polygon fill=\"#e1cff9\" points=\"2017.5,-1139 1883.5,-1139 1883.5,-1056 2017.5,-1056 2017.5,-1139\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1918\" y=\"-1123.8\">x_1 ≤ -0.24</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1905.5\" y=\"-1108.8\">entropy = 1.104</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1909.5\" y=\"-1093.8\">samples = 414</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1891.5\" y=\"-1078.8\">value = [8, 174, 232]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1915.5\" y=\"-1063.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 38&#45;&gt;39 -->\n",
       "<g class=\"edge\" id=\"edge39\">\n",
       "<title>38-&gt;39</title>\n",
       "<path d=\"M1950.5,-1174.8796C1950.5,-1166.6838 1950.5,-1157.9891 1950.5,-1149.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1954.0001,-1149.298 1950.5,-1139.2981 1947.0001,-1149.2981 1954.0001,-1149.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 110 -->\n",
       "<g class=\"node\" id=\"node111\">\n",
       "<title>110</title>\n",
       "<polygon fill=\"#fae9dc\" points=\"3266,-1139 3125,-1139 3125,-1056 3266,-1056 3266,-1139\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3159.5\" y=\"-1123.8\">x_2 ≤ -0.389</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3157.5\" y=\"-1108.8\">entropy = 1.5</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3154.5\" y=\"-1093.8\">samples = 305</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3133\" y=\"-1078.8\">value = [141, 106, 58]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3165.5\" y=\"-1063.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 38&#45;&gt;110 -->\n",
       "<g class=\"edge\" id=\"edge110\">\n",
       "<title>38-&gt;110</title>\n",
       "<path d=\"M2024.652,-1209.4124C2243.7664,-1188.4689 2887.1337,-1126.9744 3114.6486,-1105.228\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3115.2195,-1108.6894 3124.8411,-1104.2537 3114.5534,-1101.7212 3115.2195,-1108.6894\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 40 -->\n",
       "<g class=\"node\" id=\"node41\">\n",
       "<title>40</title>\n",
       "<polygon fill=\"#88efb3\" points=\"1910.5,-1020 1790.5,-1020 1790.5,-937 1910.5,-937 1910.5,-1020\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1816.5\" y=\"-1004.8\">x_2 ≤ 0.357</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1805.5\" y=\"-989.8\">entropy = 1.143</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1813\" y=\"-974.8\">samples = 88</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1798.5\" y=\"-959.8\">value = [7, 61, 20]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1813.5\" y=\"-944.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 39&#45;&gt;40 -->\n",
       "<g class=\"edge\" id=\"edge40\">\n",
       "<title>39-&gt;40</title>\n",
       "<path d=\"M1915.5249,-1055.8796C1907.9565,-1046.8733 1899.8819,-1037.2644 1892.088,-1027.9897\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1894.7374,-1025.7022 1885.6244,-1020.2981 1889.3784,-1030.2056 1894.7374,-1025.7022\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 55 -->\n",
       "<g class=\"node\" id=\"node56\">\n",
       "<title>55</title>\n",
       "<polygon fill=\"#c4a3f3\" points=\"2189.5,-1020 2055.5,-1020 2055.5,-937 2189.5,-937 2189.5,-1020\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2088.5\" y=\"-1004.8\">x_1 ≤ 0.173</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2077.5\" y=\"-989.8\">entropy = 0.959</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2081.5\" y=\"-974.8\">samples = 326</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2063.5\" y=\"-959.8\">value = [1, 113, 212]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2087.5\" y=\"-944.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 39&#45;&gt;55 -->\n",
       "<g class=\"edge\" id=\"edge55\">\n",
       "<title>39-&gt;55</title>\n",
       "<path d=\"M2010.6573,-1055.8796C2024.667,-1046.1868 2039.6854,-1035.7961 2054.025,-1025.8752\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2056.1444,-1028.6649 2062.3767,-1020.0969 2052.1617,-1022.9083 2056.1444,-1028.6649\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 41 -->\n",
       "<g class=\"node\" id=\"node42\">\n",
       "<title>41</title>\n",
       "<polygon fill=\"#a3f3c4\" points=\"1720.5,-901 1600.5,-901 1600.5,-818 1720.5,-818 1720.5,-901\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1624.5\" y=\"-885.8\">x_2 ≤ -0.162</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1619\" y=\"-870.8\">entropy = 1.19</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1623\" y=\"-855.8\">samples = 65</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1608.5\" y=\"-840.8\">value = [4, 41, 20]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1623.5\" y=\"-825.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 40&#45;&gt;41 -->\n",
       "<g class=\"edge\" id=\"edge41\">\n",
       "<title>40-&gt;41</title>\n",
       "<path d=\"M1790.3828,-940.8477C1770.8814,-928.6336 1749.1025,-914.9932 1729.0356,-902.4249\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1730.8848,-899.4533 1720.552,-897.1115 1727.1691,-905.3858 1730.8848,-899.4533\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 52 -->\n",
       "<g class=\"node\" id=\"node53\">\n",
       "<title>52</title>\n",
       "<polygon fill=\"#57e994\" points=\"1907,-901 1794,-901 1794,-818 1907,-818 1907,-901\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1816.5\" y=\"-885.8\">x_2 ≤ 0.597</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1805.5\" y=\"-870.8\">entropy = 0.559</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1813\" y=\"-855.8\">samples = 23</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1802\" y=\"-840.8\">value = [3, 20, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1813.5\" y=\"-825.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 40&#45;&gt;52 -->\n",
       "<g class=\"edge\" id=\"edge52\">\n",
       "<title>40-&gt;52</title>\n",
       "<path d=\"M1850.5,-936.8796C1850.5,-928.6838 1850.5,-919.9891 1850.5,-911.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1854.0001,-911.298 1850.5,-901.2981 1847.0001,-911.2981 1854.0001,-911.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 42 -->\n",
       "<g class=\"node\" id=\"node43\">\n",
       "<title>42</title>\n",
       "<polygon fill=\"#62ea9b\" points=\"1586,-782 1473,-782 1473,-699 1586,-699 1586,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1493.5\" y=\"-766.8\">x_1 ≤ -0.375</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1484.5\" y=\"-751.8\">entropy = 0.879</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1492\" y=\"-736.8\">samples = 43</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1481\" y=\"-721.8\">value = [4, 35, 4]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1492.5\" y=\"-706.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 41&#45;&gt;42 -->\n",
       "<g class=\"edge\" id=\"edge42\">\n",
       "<title>41-&gt;42</title>\n",
       "<path d=\"M1614.6826,-817.8796C1604.4706,-808.6031 1593.555,-798.6874 1583.0624,-789.1559\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1585.2684,-786.4313 1575.513,-782.2981 1580.5616,-791.6127 1585.2684,-786.4313\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 49 -->\n",
       "<g class=\"node\" id=\"node50\">\n",
       "<title>49</title>\n",
       "<polygon fill=\"#b083ef\" points=\"1717,-782 1604,-782 1604,-699 1717,-699 1717,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1624.5\" y=\"-766.8\">x_2 ≤ -0.026</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1615.5\" y=\"-751.8\">entropy = 0.845</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1623\" y=\"-736.8\">samples = 22</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1612\" y=\"-721.8\">value = [0, 6, 16]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1625.5\" y=\"-706.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 41&#45;&gt;49 -->\n",
       "<g class=\"edge\" id=\"edge49\">\n",
       "<title>41-&gt;49</title>\n",
       "<path d=\"M1660.5,-817.8796C1660.5,-809.6838 1660.5,-800.9891 1660.5,-792.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1664.0001,-792.298 1660.5,-782.2981 1657.0001,-792.2981 1664.0001,-792.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 43 -->\n",
       "<g class=\"node\" id=\"node44\">\n",
       "<title>43</title>\n",
       "<polygon fill=\"#9cf2c0\" points=\"1457.5,-655.5 1351.5,-655.5 1351.5,-587.5 1457.5,-587.5 1457.5,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1359.5\" y=\"-640.3\">entropy = 0.918</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1367\" y=\"-625.3\">samples = 12</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1359.5\" y=\"-610.3\">value = [4, 8, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1367.5\" y=\"-595.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 42&#45;&gt;43 -->\n",
       "<g class=\"edge\" id=\"edge43\">\n",
       "<title>42-&gt;43</title>\n",
       "<path d=\"M1485.7811,-698.8796C1473.5375,-687.2237 1460.2339,-674.5587 1448.0029,-662.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1450.2012,-660.1751 1440.5451,-655.8149 1445.3746,-665.245 1450.2012,-660.1751\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 44 -->\n",
       "<g class=\"node\" id=\"node45\">\n",
       "<title>44</title>\n",
       "<polygon fill=\"#56e994\" points=\"1589,-663 1476,-663 1476,-580 1589,-580 1589,-663\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1496.5\" y=\"-647.8\">x_2 ≤ -0.492</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1487.5\" y=\"-632.8\">entropy = 0.555</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1495\" y=\"-617.8\">samples = 31</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1484\" y=\"-602.8\">value = [0, 27, 4]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1495.5\" y=\"-587.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 42&#45;&gt;44 -->\n",
       "<g class=\"edge\" id=\"edge44\">\n",
       "<title>42-&gt;44</title>\n",
       "<path d=\"M1530.5493,-698.8796C1530.7559,-690.6838 1530.9751,-681.9891 1531.189,-673.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1534.693,-673.3831 1531.4463,-663.2981 1527.6953,-673.2067 1534.693,-673.3831\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 45 -->\n",
       "<g class=\"node\" id=\"node46\">\n",
       "<title>45</title>\n",
       "<polygon fill=\"#6eeca3\" points=\"1528,-544 1415,-544 1415,-461 1528,-461 1528,-544\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1435.5\" y=\"-528.8\">x_1 ≤ -0.311</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1426.5\" y=\"-513.8\">entropy = 0.742</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1434\" y=\"-498.8\">samples = 19</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1423\" y=\"-483.8\">value = [0, 15, 4]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1434.5\" y=\"-468.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 44&#45;&gt;45 -->\n",
       "<g class=\"edge\" id=\"edge45\">\n",
       "<title>44-&gt;45</title>\n",
       "<path d=\"M1511.1652,-579.8796C1506.7793,-571.3236 1502.1147,-562.2238 1497.583,-553.3833\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1500.6022,-551.6004 1492.9259,-544.2981 1494.373,-554.7936 1500.6022,-551.6004\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 48 -->\n",
       "<g class=\"node\" id=\"node49\">\n",
       "<title>48</title>\n",
       "<polygon fill=\"#39e581\" points=\"1659,-536.5 1546,-536.5 1546,-468.5 1659,-468.5 1659,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1564.5\" y=\"-521.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1565\" y=\"-506.3\">samples = 12</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1554\" y=\"-491.3\">value = [0, 12, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1565.5\" y=\"-476.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 44&#45;&gt;48 -->\n",
       "<g class=\"edge\" id=\"edge48\">\n",
       "<title>44-&gt;48</title>\n",
       "<path d=\"M1556.9826,-579.8796C1563.5156,-568.7735 1570.5875,-556.7513 1577.1652,-545.5691\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1580.2613,-547.2089 1582.3147,-536.8149 1574.2277,-543.6597 1580.2613,-547.2089\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 46 -->\n",
       "<g class=\"node\" id=\"node47\">\n",
       "<title>46</title>\n",
       "<polygon fill=\"#55e993\" points=\"1512.5,-417.5 1406.5,-417.5 1406.5,-349.5 1512.5,-349.5 1512.5,-417.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1414.5\" y=\"-402.3\">entropy = 0.544</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1425.5\" y=\"-387.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1414.5\" y=\"-372.3\">value = [0, 7, 1]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1422.5\" y=\"-357.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 45&#45;&gt;46 -->\n",
       "<g class=\"edge\" id=\"edge46\">\n",
       "<title>45-&gt;46</title>\n",
       "<path d=\"M1467.303,-460.8796C1466.2274,-450.2134 1465.0666,-438.7021 1463.9775,-427.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1467.4461,-427.4133 1462.9603,-417.8149 1460.4814,-428.1157 1467.4461,-427.4133\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 47 -->\n",
       "<g class=\"node\" id=\"node48\">\n",
       "<title>47</title>\n",
       "<polygon fill=\"#83efb0\" points=\"1636.5,-417.5 1530.5,-417.5 1530.5,-349.5 1636.5,-349.5 1636.5,-417.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1538.5\" y=\"-402.3\">entropy = 0.845</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1546\" y=\"-387.3\">samples = 11</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1538.5\" y=\"-372.3\">value = [0, 8, 3]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1546.5\" y=\"-357.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 45&#45;&gt;47 -->\n",
       "<g class=\"edge\" id=\"edge47\">\n",
       "<title>45-&gt;47</title>\n",
       "<path d=\"M1510.6722,-460.8796C1521.5389,-449.3337 1533.3376,-436.7976 1544.2109,-425.2446\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1546.8986,-427.4958 1551.2036,-417.8149 1541.8012,-422.6982 1546.8986,-427.4958\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 50 -->\n",
       "<g class=\"node\" id=\"node51\">\n",
       "<title>50</title>\n",
       "<polygon fill=\"#8139e5\" points=\"1713.5,-655.5 1607.5,-655.5 1607.5,-587.5 1713.5,-587.5 1713.5,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1622.5\" y=\"-640.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1626.5\" y=\"-625.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1615.5\" y=\"-610.3\">value = [0, 0, 8]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1625.5\" y=\"-595.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 49&#45;&gt;50 -->\n",
       "<g class=\"edge\" id=\"edge50\">\n",
       "<title>49-&gt;50</title>\n",
       "<path d=\"M1660.5,-698.8796C1660.5,-688.2134 1660.5,-676.7021 1660.5,-665.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1664.0001,-665.8149 1660.5,-655.8149 1657.0001,-665.815 1664.0001,-665.8149\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 51 -->\n",
       "<g class=\"node\" id=\"node52\">\n",
       "<title>51</title>\n",
       "<polygon fill=\"#e0cef8\" points=\"1837.5,-655.5 1731.5,-655.5 1731.5,-587.5 1837.5,-587.5 1837.5,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1739.5\" y=\"-640.3\">entropy = 0.985</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1747\" y=\"-625.3\">samples = 14</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1739.5\" y=\"-610.3\">value = [0, 6, 8]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1749.5\" y=\"-595.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 49&#45;&gt;51 -->\n",
       "<g class=\"edge\" id=\"edge51\">\n",
       "<title>49-&gt;51</title>\n",
       "<path d=\"M1703.8692,-698.8796C1716.0148,-687.2237 1729.2119,-674.5587 1741.3451,-662.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1743.9516,-665.2643 1748.7433,-655.8149 1739.1047,-660.2138 1743.9516,-665.2643\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 53 -->\n",
       "<g class=\"node\" id=\"node54\">\n",
       "<title>53</title>\n",
       "<polygon fill=\"#7beeab\" points=\"1841.5,-774.5 1735.5,-774.5 1735.5,-706.5 1841.5,-706.5 1841.5,-774.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1743.5\" y=\"-759.3\">entropy = 0.811</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1754.5\" y=\"-744.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1743.5\" y=\"-729.3\">value = [2, 6, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1751.5\" y=\"-714.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 52&#45;&gt;53 -->\n",
       "<g class=\"edge\" id=\"edge53\">\n",
       "<title>52-&gt;53</title>\n",
       "<path d=\"M1828.8154,-817.8796C1823.0864,-806.8835 1816.8894,-794.9893 1811.1126,-783.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1814.103,-782.0662 1806.3784,-774.8149 1807.895,-785.3007 1814.103,-782.0662\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 54 -->\n",
       "<g class=\"node\" id=\"node55\">\n",
       "<title>54</title>\n",
       "<polygon fill=\"#47e78a\" points=\"1973,-774.5 1860,-774.5 1860,-706.5 1973,-706.5 1973,-774.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1871.5\" y=\"-759.3\">entropy = 0.353</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1879\" y=\"-744.3\">samples = 15</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1868\" y=\"-729.3\">value = [1, 14, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1879.5\" y=\"-714.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 52&#45;&gt;54 -->\n",
       "<g class=\"edge\" id=\"edge54\">\n",
       "<title>52-&gt;54</title>\n",
       "<path d=\"M1873.5836,-817.8796C1879.7433,-806.7735 1886.4111,-794.7513 1892.6129,-783.5691\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1895.6787,-785.2576 1897.4682,-774.8149 1889.5572,-781.8624 1895.6787,-785.2576\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 56 -->\n",
       "<g class=\"node\" id=\"node57\">\n",
       "<title>56</title>\n",
       "<polygon fill=\"#9d65eb\" points=\"2186,-901 2059,-901 2059,-818 2186,-818 2186,-901\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2086.5\" y=\"-885.8\">x_1 ≤ -0.175</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2077.5\" y=\"-870.8\">entropy = 0.688</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2081.5\" y=\"-855.8\">samples = 218</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2067\" y=\"-840.8\">value = [0, 40, 178]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2087.5\" y=\"-825.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 55&#45;&gt;56 -->\n",
       "<g class=\"edge\" id=\"edge56\">\n",
       "<title>55-&gt;56</title>\n",
       "<path d=\"M2122.5,-936.8796C2122.5,-928.6838 2122.5,-919.9891 2122.5,-911.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2126.0001,-911.298 2122.5,-901.2981 2119.0001,-911.2981 2126.0001,-911.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 91 -->\n",
       "<g class=\"node\" id=\"node92\">\n",
       "<title>91</title>\n",
       "<polygon fill=\"#97f1bd\" points=\"2603.5,-901 2483.5,-901 2483.5,-818 2603.5,-818 2603.5,-901\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2509.5\" y=\"-885.8\">x_1 ≤ 0.303</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2498.5\" y=\"-870.8\">entropy = 0.969</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2502.5\" y=\"-855.8\">samples = 108</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2491.5\" y=\"-840.8\">value = [1, 73, 34]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2506.5\" y=\"-825.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 55&#45;&gt;91 -->\n",
       "<g class=\"edge\" id=\"edge91\">\n",
       "<title>55-&gt;91</title>\n",
       "<path d=\"M2189.6742,-959.5125C2267.3708,-937.5508 2394.6797,-901.5656 2473.5396,-879.275\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2474.7137,-882.5804 2483.3846,-876.4922 2472.8096,-875.8443 2474.7137,-882.5804\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 57 -->\n",
       "<g class=\"node\" id=\"node58\">\n",
       "<title>57</title>\n",
       "<polygon fill=\"#f7f3fd\" points=\"2111.5,-782 1991.5,-782 1991.5,-699 2111.5,-699 2111.5,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2015.5\" y=\"-766.8\">x_2 ≤ -0.337</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2006.5\" y=\"-751.8\">entropy = 0.999</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2014\" y=\"-736.8\">samples = 31</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1999.5\" y=\"-721.8\">value = [0, 15, 16]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2016.5\" y=\"-706.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 56&#45;&gt;57 -->\n",
       "<g class=\"edge\" id=\"edge57\">\n",
       "<title>56-&gt;57</title>\n",
       "<path d=\"M2097.6676,-817.8796C2092.5091,-809.2335 2087.0192,-800.0322 2081.6924,-791.1042\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2084.5678,-789.0924 2076.4383,-782.2981 2078.5564,-792.679 2084.5678,-789.0924\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 62 -->\n",
       "<g class=\"node\" id=\"node63\">\n",
       "<title>62</title>\n",
       "<polygon fill=\"#9458e9\" points=\"2257,-782 2130,-782 2130,-699 2257,-699 2257,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2159.5\" y=\"-766.8\">x_2 ≤ 0.889</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2148.5\" y=\"-751.8\">entropy = 0.567</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2152.5\" y=\"-736.8\">samples = 187</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2138\" y=\"-721.8\">value = [0, 25, 162]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2158.5\" y=\"-706.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 56&#45;&gt;62 -->\n",
       "<g class=\"edge\" id=\"edge62\">\n",
       "<title>56-&gt;62</title>\n",
       "<path d=\"M2147.3324,-817.8796C2152.4909,-809.2335 2157.9808,-800.0322 2163.3076,-791.1042\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2166.4436,-792.679 2168.5617,-782.2981 2160.4322,-789.0924 2166.4436,-792.679\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 58 -->\n",
       "<g class=\"node\" id=\"node59\">\n",
       "<title>58</title>\n",
       "<polygon fill=\"#7beeab\" points=\"1968.5,-655.5 1862.5,-655.5 1862.5,-587.5 1968.5,-587.5 1968.5,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1870.5\" y=\"-640.3\">entropy = 0.811</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1881.5\" y=\"-625.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1870.5\" y=\"-610.3\">value = [0, 6, 2]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1878.5\" y=\"-595.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 57&#45;&gt;58 -->\n",
       "<g class=\"edge\" id=\"edge58\">\n",
       "<title>57-&gt;58</title>\n",
       "<path d=\"M2003.9338,-698.8796C1990.4872,-687.1138 1975.8653,-674.3197 1962.4549,-662.5855\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1964.5476,-659.766 1954.7171,-655.8149 1959.9381,-665.034 1964.5476,-659.766\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 59 -->\n",
       "<g class=\"node\" id=\"node60\">\n",
       "<title>59</title>\n",
       "<polygon fill=\"#d2b8f6\" points=\"2100,-663 1987,-663 1987,-580 2100,-580 2100,-663\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2007.5\" y=\"-647.8\">x_1 ≤ -0.193</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1998.5\" y=\"-632.8\">entropy = 0.966</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2006\" y=\"-617.8\">samples = 23</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1995\" y=\"-602.8\">value = [0, 9, 14]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2008.5\" y=\"-587.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 57&#45;&gt;59 -->\n",
       "<g class=\"edge\" id=\"edge59\">\n",
       "<title>57-&gt;59</title>\n",
       "<path d=\"M2048.702,-698.8796C2048.151,-690.6838 2047.5665,-681.9891 2046.9959,-673.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2050.4729,-673.0408 2046.31,-663.2981 2043.4887,-673.5103 2050.4729,-673.0408\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 60 -->\n",
       "<g class=\"node\" id=\"node61\">\n",
       "<title>60</title>\n",
       "<polygon fill=\"#9d65eb\" points=\"1964.5,-536.5 1858.5,-536.5 1858.5,-468.5 1964.5,-468.5 1964.5,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1866.5\" y=\"-521.3\">entropy = 0.684</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1874\" y=\"-506.3\">samples = 11</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1866.5\" y=\"-491.3\">value = [0, 2, 9]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1876.5\" y=\"-476.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 59&#45;&gt;60 -->\n",
       "<g class=\"edge\" id=\"edge60\">\n",
       "<title>59-&gt;60</title>\n",
       "<path d=\"M1997.3328,-579.8796C1984.2817,-568.1138 1970.0899,-555.3197 1957.0739,-543.5855\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1959.3346,-540.9113 1949.5636,-536.8149 1954.6474,-546.1104 1959.3346,-540.9113\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 61 -->\n",
       "<g class=\"node\" id=\"node62\">\n",
       "<title>61</title>\n",
       "<polygon fill=\"#c6f8db\" points=\"2088.5,-536.5 1982.5,-536.5 1982.5,-468.5 2088.5,-468.5 2088.5,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1994\" y=\"-521.3\">entropy = 0.98</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1998\" y=\"-506.3\">samples = 12</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1990.5\" y=\"-491.3\">value = [0, 7, 5]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1998.5\" y=\"-476.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 59&#45;&gt;61 -->\n",
       "<g class=\"edge\" id=\"edge61\">\n",
       "<title>59-&gt;61</title>\n",
       "<path d=\"M2040.702,-579.8796C2039.9849,-569.2134 2039.2111,-557.7021 2038.485,-546.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2041.9699,-546.5576 2037.8069,-536.8149 2034.9856,-547.0272 2041.9699,-546.5576\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 63 -->\n",
       "<g class=\"node\" id=\"node64\">\n",
       "<title>63</title>\n",
       "<polygon fill=\"#965ae9\" points=\"2251,-663 2124,-663 2124,-580 2251,-580 2251,-663\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2153.5\" y=\"-647.8\">x_1 ≤ 0.059</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2142.5\" y=\"-632.8\">entropy = 0.596</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2146.5\" y=\"-617.8\">samples = 173</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2132\" y=\"-602.8\">value = [0, 25, 148]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2152.5\" y=\"-587.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 62&#45;&gt;63 -->\n",
       "<g class=\"edge\" id=\"edge63\">\n",
       "<title>62-&gt;63</title>\n",
       "<path d=\"M2191.4015,-698.8796C2190.9883,-690.6838 2190.5499,-681.9891 2190.1219,-673.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2193.6067,-673.1091 2189.6075,-663.2981 2186.6156,-673.4617 2193.6067,-673.1091\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 90 -->\n",
       "<g class=\"node\" id=\"node91\">\n",
       "<title>90</title>\n",
       "<polygon fill=\"#8139e5\" points=\"2382,-655.5 2269,-655.5 2269,-587.5 2382,-587.5 2382,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2287.5\" y=\"-640.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2288\" y=\"-625.3\">samples = 14</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2277\" y=\"-610.3\">value = [0, 0, 14]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2290.5\" y=\"-595.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 62&#45;&gt;90 -->\n",
       "<g class=\"edge\" id=\"edge90\">\n",
       "<title>62-&gt;90</title>\n",
       "<path d=\"M2239.6672,-698.8796C2252.7183,-687.1138 2266.9101,-674.3197 2279.9261,-662.5855\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2282.3526,-665.1104 2287.4364,-655.8149 2277.6654,-659.9113 2282.3526,-665.1104\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 64 -->\n",
       "<g class=\"node\" id=\"node65\">\n",
       "<title>64</title>\n",
       "<polygon fill=\"#9050e8\" points=\"2234,-544 2107,-544 2107,-461 2234,-461 2234,-544\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2136.5\" y=\"-528.8\">x_2 ≤ 0.392</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2125.5\" y=\"-513.8\">entropy = 0.485</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2129.5\" y=\"-498.8\">samples = 114</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2115\" y=\"-483.8\">value = [0, 12, 102]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2135.5\" y=\"-468.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 63&#45;&gt;64 -->\n",
       "<g class=\"edge\" id=\"edge64\">\n",
       "<title>63-&gt;64</title>\n",
       "<path d=\"M2181.5542,-579.8796C2180.3705,-571.5938 2179.114,-562.798 2177.8888,-554.2216\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2181.3503,-553.7026 2176.4712,-544.2981 2174.4206,-554.6926 2181.3503,-553.7026\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 79 -->\n",
       "<g class=\"node\" id=\"node80\">\n",
       "<title>79</title>\n",
       "<polygon fill=\"#a571ec\" points=\"2372.5,-544 2252.5,-544 2252.5,-461 2372.5,-461 2372.5,-544\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2285.5\" y=\"-528.8\">x_1 ≤ 0.1</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2267.5\" y=\"-513.8\">entropy = 0.761</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2275\" y=\"-498.8\">samples = 59</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2260.5\" y=\"-483.8\">value = [0, 13, 46]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2277.5\" y=\"-468.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 63&#45;&gt;79 -->\n",
       "<g class=\"edge\" id=\"edge79\">\n",
       "<title>63-&gt;79</title>\n",
       "<path d=\"M2231.2189,-579.8796C2240.8686,-570.6931 2251.1766,-560.8798 2261.0991,-551.4336\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2263.765,-553.7282 2268.5945,-544.2981 2258.9384,-548.6582 2263.765,-553.7282\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 65 -->\n",
       "<g class=\"node\" id=\"node66\">\n",
       "<title>65</title>\n",
       "<polygon fill=\"#8b48e7\" points=\"1776,-425 1663,-425 1663,-342 1776,-342 1776,-425\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1683.5\" y=\"-409.8\">x_2 ≤ -0.314</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1674.5\" y=\"-394.8\">entropy = 0.368</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1682\" y=\"-379.8\">samples = 85</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1671\" y=\"-364.8\">value = [0, 6, 79]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1684.5\" y=\"-349.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 64&#45;&gt;65 -->\n",
       "<g class=\"edge\" id=\"edge65\">\n",
       "<title>64-&gt;65</title>\n",
       "<path d=\"M2106.9807,-464.826C2103.8163,-463.4454 2100.6462,-462.1594 2097.5,-461 2017.6739,-431.5836 1992.8418,-442.0784 1909.5,-425 1868.533,-416.6051 1822.7327,-406.6192 1786.3652,-398.5469\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1786.8911,-395.0784 1776.3697,-396.3238 1785.3713,-401.9115 1786.8911,-395.0784\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 74 -->\n",
       "<g class=\"node\" id=\"node75\">\n",
       "<title>74</title>\n",
       "<polygon fill=\"#a26dec\" points=\"2032,-425 1919,-425 1919,-342 2032,-342 2032,-425\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1939.5\" y=\"-409.8\">x_1 ≤ -0.119</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1930.5\" y=\"-394.8\">entropy = 0.736</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1938\" y=\"-379.8\">samples = 29</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1927\" y=\"-364.8\">value = [0, 6, 23]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1940.5\" y=\"-349.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 64&#45;&gt;74 -->\n",
       "<g class=\"edge\" id=\"edge74\">\n",
       "<title>64-&gt;74</title>\n",
       "<path d=\"M2106.9195,-463.6996C2085.7222,-450.7638 2062.0813,-436.3368 2040.7587,-423.3245\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2042.4491,-420.2559 2032.0898,-418.0343 2038.8026,-426.2312 2042.4491,-420.2559\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 66 -->\n",
       "<g class=\"node\" id=\"node67\">\n",
       "<title>66</title>\n",
       "<polygon fill=\"#9356e9\" points=\"1645,-306 1532,-306 1532,-223 1645,-223 1645,-306\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1552.5\" y=\"-290.8\">x_1 ≤ -0.075</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1543.5\" y=\"-275.8\">entropy = 0.551</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1551\" y=\"-260.8\">samples = 47</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1540\" y=\"-245.8\">value = [0, 6, 41]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1553.5\" y=\"-230.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 65&#45;&gt;66 -->\n",
       "<g class=\"edge\" id=\"edge66\">\n",
       "<title>65-&gt;66</title>\n",
       "<path d=\"M1673.6826,-341.8796C1663.4706,-332.6031 1652.555,-322.6874 1642.0624,-313.1559\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1644.2684,-310.4313 1634.513,-306.2981 1639.5616,-315.6127 1644.2684,-310.4313\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 73 -->\n",
       "<g class=\"node\" id=\"node74\">\n",
       "<title>73</title>\n",
       "<polygon fill=\"#8139e5\" points=\"1776,-298.5 1663,-298.5 1663,-230.5 1776,-230.5 1776,-298.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1681.5\" y=\"-283.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1682\" y=\"-268.3\">samples = 38</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1671\" y=\"-253.3\">value = [0, 0, 38]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1684.5\" y=\"-238.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 65&#45;&gt;73 -->\n",
       "<g class=\"edge\" id=\"edge73\">\n",
       "<title>65-&gt;73</title>\n",
       "<path d=\"M1719.5,-341.8796C1719.5,-331.2134 1719.5,-319.7021 1719.5,-308.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1723.0001,-308.8149 1719.5,-298.8149 1716.0001,-308.815 1723.0001,-308.8149\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 67 -->\n",
       "<g class=\"node\" id=\"node68\">\n",
       "<title>67</title>\n",
       "<polygon fill=\"#ab7bee\" points=\"1580,-187 1467,-187 1467,-104 1580,-104 1580,-187\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1487.5\" y=\"-171.8\">x_1 ≤ -0.122</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1478.5\" y=\"-156.8\">entropy = 0.811</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1486\" y=\"-141.8\">samples = 16</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1475\" y=\"-126.8\">value = [0, 4, 12]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1488.5\" y=\"-111.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 66&#45;&gt;67 -->\n",
       "<g class=\"edge\" id=\"edge67\">\n",
       "<title>66-&gt;67</title>\n",
       "<path d=\"M1565.7662,-222.8796C1561.0435,-214.2335 1556.0176,-205.0322 1551.141,-196.1042\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1554.1963,-194.3964 1546.3309,-187.2981 1548.053,-197.752 1554.1963,-194.3964\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 70 -->\n",
       "<g class=\"node\" id=\"node71\">\n",
       "<title>70</title>\n",
       "<polygon fill=\"#8a47e7\" points=\"1711,-187 1598,-187 1598,-104 1711,-104 1711,-187\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1618.5\" y=\"-171.8\">x_1 ≤ -0.006</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1609.5\" y=\"-156.8\">entropy = 0.345</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1617\" y=\"-141.8\">samples = 31</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1606\" y=\"-126.8\">value = [0, 2, 29]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1619.5\" y=\"-111.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 66&#45;&gt;70 -->\n",
       "<g class=\"edge\" id=\"edge70\">\n",
       "<title>66-&gt;70</title>\n",
       "<path d=\"M1611.5836,-222.8796C1616.3789,-214.2335 1621.4822,-205.0322 1626.4338,-196.1042\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1629.5284,-197.7407 1631.3179,-187.2981 1623.4069,-194.3455 1629.5284,-197.7407\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 68 -->\n",
       "<g class=\"node\" id=\"node69\">\n",
       "<title>68</title>\n",
       "<polygon fill=\"#ab7bee\" points=\"1453.5,-68 1347.5,-68 1347.5,0 1453.5,0 1453.5,-68\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1355.5\" y=\"-52.8\">entropy = 0.811</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1366.5\" y=\"-37.8\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1355.5\" y=\"-22.8\">value = [0, 2, 6]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1365.5\" y=\"-7.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 67&#45;&gt;68 -->\n",
       "<g class=\"edge\" id=\"edge68\">\n",
       "<title>67-&gt;68</title>\n",
       "<path d=\"M1477.6993,-103.9815C1467.2566,-94.5151 1456.1667,-84.462 1445.7472,-75.0168\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1448.0572,-72.3868 1438.2976,-68.2637 1443.3559,-77.5731 1448.0572,-72.3868\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 69 -->\n",
       "<g class=\"node\" id=\"node70\">\n",
       "<title>69</title>\n",
       "<polygon fill=\"#ab7bee\" points=\"1577.5,-68 1471.5,-68 1471.5,0 1577.5,0 1577.5,-68\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1479.5\" y=\"-52.8\">entropy = 0.811</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1490.5\" y=\"-37.8\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1479.5\" y=\"-22.8\">value = [0, 2, 6]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1489.5\" y=\"-7.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 67&#45;&gt;69 -->\n",
       "<g class=\"edge\" id=\"edge69\">\n",
       "<title>67-&gt;69</title>\n",
       "<path d=\"M1523.8724,-103.9815C1523.9474,-95.618 1524.0265,-86.7965 1524.1023,-78.3409\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1527.6028,-78.2947 1524.1927,-68.2637 1520.6031,-78.2318 1527.6028,-78.2947\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 71 -->\n",
       "<g class=\"node\" id=\"node72\">\n",
       "<title>71</title>\n",
       "<polygon fill=\"#8139e5\" points=\"1710,-68 1597,-68 1597,0 1710,0 1710,-68\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1615.5\" y=\"-52.8\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1616\" y=\"-37.8\">samples = 16</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1605\" y=\"-22.8\">value = [0, 0, 16]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1618.5\" y=\"-7.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 70&#45;&gt;71 -->\n",
       "<g class=\"edge\" id=\"edge71\">\n",
       "<title>70-&gt;71</title>\n",
       "<path d=\"M1654.1276,-103.9815C1654.0526,-95.618 1653.9735,-86.7965 1653.8977,-78.3409\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1657.3969,-78.2318 1653.8073,-68.2637 1650.3972,-78.2947 1657.3969,-78.2318\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 72 -->\n",
       "<g class=\"node\" id=\"node73\">\n",
       "<title>72</title>\n",
       "<polygon fill=\"#9457e9\" points=\"1841,-68 1728,-68 1728,0 1841,0 1841,-68\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1739.5\" y=\"-52.8\">entropy = 0.567</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1747\" y=\"-37.8\">samples = 15</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1736\" y=\"-22.8\">value = [0, 2, 13]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1749.5\" y=\"-7.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 70&#45;&gt;72 -->\n",
       "<g class=\"edge\" id=\"edge72\">\n",
       "<title>70-&gt;72</title>\n",
       "<path d=\"M1702.9072,-103.9815C1713.9442,-94.5151 1725.6653,-84.462 1736.6777,-75.0168\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1739.2394,-77.4307 1744.5513,-68.2637 1734.6822,-72.1173 1739.2394,-77.4307\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 75 -->\n",
       "<g class=\"node\" id=\"node76\">\n",
       "<title>75</title>\n",
       "<polygon fill=\"#c09cf2\" points=\"1900.5,-298.5 1794.5,-298.5 1794.5,-230.5 1900.5,-230.5 1900.5,-298.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1802.5\" y=\"-283.3\">entropy = 0.918</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1813.5\" y=\"-268.3\">samples = 9</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1802.5\" y=\"-253.3\">value = [0, 3, 6]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1812.5\" y=\"-238.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 74&#45;&gt;75 -->\n",
       "<g class=\"edge\" id=\"edge75\">\n",
       "<title>74-&gt;75</title>\n",
       "<path d=\"M1930.7318,-341.8796C1918.1944,-330.2237 1904.5716,-317.5587 1892.047,-305.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1894.1172,-303.0605 1884.4102,-298.8149 1889.3509,-308.1872 1894.1172,-303.0605\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 76 -->\n",
       "<g class=\"node\" id=\"node77\">\n",
       "<title>76</title>\n",
       "<polygon fill=\"#975cea\" points=\"2032,-306 1919,-306 1919,-223 2032,-223 2032,-306\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1945\" y=\"-290.8\">x_2 ≤ 0.64</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1934\" y=\"-275.8\">entropy = 0.61</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1938\" y=\"-260.8\">samples = 20</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1927\" y=\"-245.8\">value = [0, 3, 17]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1940.5\" y=\"-230.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 74&#45;&gt;76 -->\n",
       "<g class=\"edge\" id=\"edge76\">\n",
       "<title>74-&gt;76</title>\n",
       "<path d=\"M1975.5,-341.8796C1975.5,-333.6838 1975.5,-324.9891 1975.5,-316.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1979.0001,-316.298 1975.5,-306.2981 1972.0001,-316.2981 1979.0001,-316.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 77 -->\n",
       "<g class=\"node\" id=\"node78\">\n",
       "<title>77</title>\n",
       "<polygon fill=\"#b083ef\" points=\"1968.5,-179.5 1862.5,-179.5 1862.5,-111.5 1968.5,-111.5 1968.5,-179.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1870.5\" y=\"-164.3\">entropy = 0.845</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1878\" y=\"-149.3\">samples = 11</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1870.5\" y=\"-134.3\">value = [0, 3, 8]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1880.5\" y=\"-119.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 76&#45;&gt;77 -->\n",
       "<g class=\"edge\" id=\"edge77\">\n",
       "<title>76-&gt;77</title>\n",
       "<path d=\"M1954.5149,-222.8796C1948.9707,-211.8835 1942.9736,-199.9893 1937.3831,-188.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"1940.4291,-187.1684 1932.8016,-179.8149 1934.1786,-190.3199 1940.4291,-187.1684\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 78 -->\n",
       "<g class=\"node\" id=\"node79\">\n",
       "<title>78</title>\n",
       "<polygon fill=\"#8139e5\" points=\"2092.5,-179.5 1986.5,-179.5 1986.5,-111.5 2092.5,-111.5 2092.5,-179.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2001.5\" y=\"-164.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2005.5\" y=\"-149.3\">samples = 9</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"1994.5\" y=\"-134.3\">value = [0, 0, 9]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2004.5\" y=\"-119.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 76&#45;&gt;78 -->\n",
       "<g class=\"edge\" id=\"edge78\">\n",
       "<title>76-&gt;78</title>\n",
       "<path d=\"M1997.8841,-222.8796C2003.798,-211.8835 2010.1949,-199.9893 2016.158,-188.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2019.3908,-190.2799 2021.0449,-179.8149 2013.2258,-186.9642 2019.3908,-190.2799\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 80 -->\n",
       "<g class=\"node\" id=\"node81\">\n",
       "<title>80</title>\n",
       "<polygon fill=\"#d1b7f6\" points=\"2282,-425 2169,-425 2169,-342 2282,-342 2282,-425\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2191.5\" y=\"-409.8\">x_1 ≤ 0.087</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2180.5\" y=\"-394.8\">entropy = 0.964</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2188\" y=\"-379.8\">samples = 18</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2177\" y=\"-364.8\">value = [0, 7, 11]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2190.5\" y=\"-349.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 79&#45;&gt;80 -->\n",
       "<g class=\"edge\" id=\"edge80\">\n",
       "<title>79-&gt;80</title>\n",
       "<path d=\"M2282.0716,-460.8796C2275.6189,-452.0534 2268.743,-442.6485 2262.0887,-433.5466\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2264.7856,-431.3051 2256.0583,-425.2981 2259.1347,-435.4364 2264.7856,-431.3051\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 83 -->\n",
       "<g class=\"node\" id=\"node84\">\n",
       "<title>83</title>\n",
       "<polygon fill=\"#975be9\" points=\"2413,-425 2300,-425 2300,-342 2413,-342 2413,-425\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2322.5\" y=\"-409.8\">x_1 ≤ 0.136</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2311.5\" y=\"-394.8\">entropy = 0.601</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2319\" y=\"-379.8\">samples = 41</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2308\" y=\"-364.8\">value = [0, 6, 35]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2321.5\" y=\"-349.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 79&#45;&gt;83 -->\n",
       "<g class=\"edge\" id=\"edge83\">\n",
       "<title>79-&gt;83</title>\n",
       "<path d=\"M2327.8891,-460.8796C2330.986,-452.5037 2334.2757,-443.6067 2337.4794,-434.942\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2340.86,-435.8913 2341.0453,-425.2981 2334.2944,-433.4636 2340.86,-435.8913\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 81 -->\n",
       "<g class=\"node\" id=\"node82\">\n",
       "<title>81</title>\n",
       "<polygon fill=\"#a06aec\" points=\"2156.5,-298.5 2050.5,-298.5 2050.5,-230.5 2156.5,-230.5 2156.5,-298.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2058.5\" y=\"-283.3\">entropy = 0.722</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2066\" y=\"-268.3\">samples = 10</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2058.5\" y=\"-253.3\">value = [0, 2, 8]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2068.5\" y=\"-238.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 80&#45;&gt;81 -->\n",
       "<g class=\"edge\" id=\"edge81\">\n",
       "<title>80-&gt;81</title>\n",
       "<path d=\"M2182.8303,-341.8796C2170.8806,-330.2237 2157.8963,-317.5587 2145.9589,-305.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2148.2825,-303.292 2138.68,-298.8149 2143.3947,-308.303 2148.2825,-303.292\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 82 -->\n",
       "<g class=\"node\" id=\"node83\">\n",
       "<title>82</title>\n",
       "<polygon fill=\"#b0f5cd\" points=\"2280.5,-298.5 2174.5,-298.5 2174.5,-230.5 2280.5,-230.5 2280.5,-298.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2182.5\" y=\"-283.3\">entropy = 0.954</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2193.5\" y=\"-268.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2182.5\" y=\"-253.3\">value = [0, 5, 3]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2190.5\" y=\"-238.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 80&#45;&gt;82 -->\n",
       "<g class=\"edge\" id=\"edge82\">\n",
       "<title>80-&gt;82</title>\n",
       "<path d=\"M2226.1995,-341.8796C2226.3788,-331.2134 2226.5722,-319.7021 2226.7538,-308.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2230.2546,-308.8724 2226.9233,-298.8149 2223.2556,-308.7547 2230.2546,-308.8724\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 84 -->\n",
       "<g class=\"node\" id=\"node85\">\n",
       "<title>84</title>\n",
       "<polygon fill=\"#8844e6\" points=\"2412,-306 2299,-306 2299,-223 2412,-223 2412,-306\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2321.5\" y=\"-290.8\">x_1 ≤ 0.115</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2310.5\" y=\"-275.8\">entropy = 0.297</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2318\" y=\"-260.8\">samples = 19</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2307\" y=\"-245.8\">value = [0, 1, 18]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2320.5\" y=\"-230.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 83&#45;&gt;84 -->\n",
       "<g class=\"edge\" id=\"edge84\">\n",
       "<title>83-&gt;84</title>\n",
       "<path d=\"M2356.1502,-341.8796C2356.0814,-333.6838 2356.0083,-324.9891 2355.937,-316.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2359.4352,-316.2683 2355.8512,-306.2981 2352.4355,-316.3272 2359.4352,-316.2683\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 87 -->\n",
       "<g class=\"node\" id=\"node88\">\n",
       "<title>87</title>\n",
       "<polygon fill=\"#a673ed\" points=\"2543,-306 2430,-306 2430,-223 2543,-223 2543,-306\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2450.5\" y=\"-290.8\">x_2 ≤ -0.154</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2441.5\" y=\"-275.8\">entropy = 0.773</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2449\" y=\"-260.8\">samples = 22</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2438\" y=\"-245.8\">value = [0, 5, 17]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2451.5\" y=\"-230.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 83&#45;&gt;87 -->\n",
       "<g class=\"edge\" id=\"edge87\">\n",
       "<title>83-&gt;87</title>\n",
       "<path d=\"M2401.9677,-341.8796C2412.1017,-332.6031 2422.934,-322.6874 2433.3465,-313.1559\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2435.8252,-315.6319 2440.8382,-306.2981 2431.0987,-310.4685 2435.8252,-315.6319\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 85 -->\n",
       "<g class=\"node\" id=\"node86\">\n",
       "<title>85</title>\n",
       "<polygon fill=\"#9355e9\" points=\"2250.5,-179.5 2144.5,-179.5 2144.5,-111.5 2250.5,-111.5 2250.5,-179.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2152.5\" y=\"-164.3\">entropy = 0.544</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2163.5\" y=\"-149.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2152.5\" y=\"-134.3\">value = [0, 1, 7]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2162.5\" y=\"-119.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 84&#45;&gt;85 -->\n",
       "<g class=\"edge\" id=\"edge85\">\n",
       "<title>84-&gt;85</title>\n",
       "<path d=\"M2300.2393,-222.8796C2284.2497,-210.8368 2266.8297,-197.7167 2250.9526,-185.7586\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2252.8998,-182.8435 2242.8062,-179.623 2248.6884,-188.435 2252.8998,-182.8435\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 86 -->\n",
       "<g class=\"node\" id=\"node87\">\n",
       "<title>86</title>\n",
       "<polygon fill=\"#8139e5\" points=\"2382,-179.5 2269,-179.5 2269,-111.5 2382,-111.5 2382,-179.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2287.5\" y=\"-164.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2288\" y=\"-149.3\">samples = 11</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2277\" y=\"-134.3\">value = [0, 0, 11]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2290.5\" y=\"-119.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 84&#45;&gt;86 -->\n",
       "<g class=\"edge\" id=\"edge86\">\n",
       "<title>84-&gt;86</title>\n",
       "<path d=\"M2345.0075,-222.8796C2342.2908,-212.1034 2339.3566,-200.4647 2336.6095,-189.5677\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2339.9892,-188.6559 2334.1508,-179.8149 2333.2016,-190.3672 2339.9892,-188.6559\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 88 -->\n",
       "<g class=\"node\" id=\"node89\">\n",
       "<title>88</title>\n",
       "<polygon fill=\"#c09cf2\" points=\"2506.5,-179.5 2400.5,-179.5 2400.5,-111.5 2506.5,-111.5 2506.5,-179.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2408.5\" y=\"-164.3\">entropy = 0.918</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2416\" y=\"-149.3\">samples = 12</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2408.5\" y=\"-134.3\">value = [0, 4, 8]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2418.5\" y=\"-119.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 87&#45;&gt;88 -->\n",
       "<g class=\"edge\" id=\"edge88\">\n",
       "<title>87-&gt;88</title>\n",
       "<path d=\"M2474.9582,-222.8796C2471.9699,-212.1034 2468.7423,-200.4647 2465.7205,-189.5677\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2469.061,-188.516 2463.0159,-179.8149 2462.3155,-190.3866 2469.061,-188.516\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 89 -->\n",
       "<g class=\"node\" id=\"node90\">\n",
       "<title>89</title>\n",
       "<polygon fill=\"#8f4fe8\" points=\"2630.5,-179.5 2524.5,-179.5 2524.5,-111.5 2630.5,-111.5 2630.5,-179.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2532.5\" y=\"-164.3\">entropy = 0.469</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2540\" y=\"-149.3\">samples = 10</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2532.5\" y=\"-134.3\">value = [0, 1, 9]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2542.5\" y=\"-119.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 87&#45;&gt;89 -->\n",
       "<g class=\"edge\" id=\"edge89\">\n",
       "<title>87-&gt;89</title>\n",
       "<path d=\"M2518.3274,-222.8796C2526.9884,-211.5536 2536.3781,-199.2748 2545.072,-187.9058\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2547.9649,-189.8846 2551.2592,-179.8149 2542.4044,-185.6324 2547.9649,-189.8846\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 92 -->\n",
       "<g class=\"node\" id=\"node93\">\n",
       "<title>92</title>\n",
       "<polygon fill=\"#bff7d6\" points=\"2603.5,-782 2483.5,-782 2483.5,-699 2603.5,-699 2603.5,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2507.5\" y=\"-766.8\">x_2 ≤ -0.853</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2498.5\" y=\"-751.8\">entropy = 1.064</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2506\" y=\"-736.8\">samples = 71</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2491.5\" y=\"-721.8\">value = [1, 42, 28]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2506.5\" y=\"-706.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 91&#45;&gt;92 -->\n",
       "<g class=\"edge\" id=\"edge92\">\n",
       "<title>91-&gt;92</title>\n",
       "<path d=\"M2543.5,-817.8796C2543.5,-809.6838 2543.5,-800.9891 2543.5,-792.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2547.0001,-792.298 2543.5,-782.2981 2540.0001,-792.2981 2547.0001,-792.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 105 -->\n",
       "<g class=\"node\" id=\"node106\">\n",
       "<title>105</title>\n",
       "<polygon fill=\"#5fea99\" points=\"2800,-782 2687,-782 2687,-699 2800,-699 2800,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2709.5\" y=\"-766.8\">x_2 ≤ 0.227</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2698.5\" y=\"-751.8\">entropy = 0.639</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2706\" y=\"-736.8\">samples = 37</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2695\" y=\"-721.8\">value = [0, 31, 6]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2706.5\" y=\"-706.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 91&#45;&gt;105 -->\n",
       "<g class=\"edge\" id=\"edge105\">\n",
       "<title>91-&gt;105</title>\n",
       "<path d=\"M2603.5066,-823.7961C2627.0761,-809.7722 2654.1945,-793.6368 2678.2493,-779.3241\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2680.1566,-782.262 2686.9608,-774.1408 2676.5773,-776.2464 2680.1566,-782.262\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 93 -->\n",
       "<g class=\"node\" id=\"node94\">\n",
       "<title>93</title>\n",
       "<polygon fill=\"#d7fae6\" points=\"2530.5,-655.5 2424.5,-655.5 2424.5,-587.5 2530.5,-587.5 2530.5,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2432.5\" y=\"-640.3\">entropy = 1.406</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2443.5\" y=\"-625.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2432.5\" y=\"-610.3\">value = [1, 4, 3]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2440.5\" y=\"-595.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 92&#45;&gt;93 -->\n",
       "<g class=\"edge\" id=\"edge93\">\n",
       "<title>92-&gt;93</title>\n",
       "<path d=\"M2520.4164,-698.8796C2514.2567,-687.7735 2507.5889,-675.7513 2501.3871,-664.5691\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2504.4428,-662.8624 2496.5318,-655.8149 2498.3213,-666.2576 2504.4428,-662.8624\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 94 -->\n",
       "<g class=\"node\" id=\"node95\">\n",
       "<title>94</title>\n",
       "<polygon fill=\"#bbf6d4\" points=\"2668.5,-663 2548.5,-663 2548.5,-580 2668.5,-580 2668.5,-663\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2574.5\" y=\"-647.8\">x_1 ≤ 0.262</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2563.5\" y=\"-632.8\">entropy = 0.969</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2571\" y=\"-617.8\">samples = 63</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2556.5\" y=\"-602.8\">value = [0, 38, 25]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2571.5\" y=\"-587.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 92&#45;&gt;94 -->\n",
       "<g class=\"edge\" id=\"edge94\">\n",
       "<title>92-&gt;94</title>\n",
       "<path d=\"M2566.2338,-698.8796C2570.9565,-690.2335 2575.9824,-681.0322 2580.859,-672.1042\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2583.947,-673.752 2585.6691,-663.2981 2577.8037,-670.3964 2583.947,-673.752\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 95 -->\n",
       "<g class=\"node\" id=\"node96\">\n",
       "<title>95</title>\n",
       "<polygon fill=\"#98f2be\" points=\"2664.5,-544 2544.5,-544 2544.5,-461 2664.5,-461 2664.5,-544\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2570.5\" y=\"-528.8\">x_1 ≤ 0.232</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2563\" y=\"-513.8\">entropy = 0.91</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2567\" y=\"-498.8\">samples = 40</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2552.5\" y=\"-483.8\">value = [0, 27, 13]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2567.5\" y=\"-468.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 94&#45;&gt;95 -->\n",
       "<g class=\"edge\" id=\"edge95\">\n",
       "<title>94-&gt;95</title>\n",
       "<path d=\"M2607.101,-579.8796C2606.8255,-571.6838 2606.5332,-562.9891 2606.2479,-554.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2609.739,-554.1748 2605.905,-544.2981 2602.743,-554.41 2609.739,-554.1748\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 102 -->\n",
       "<g class=\"node\" id=\"node103\">\n",
       "<title>102</title>\n",
       "<polygon fill=\"#f5effd\" points=\"2802.5,-544 2682.5,-544 2682.5,-461 2802.5,-461 2802.5,-544\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2706.5\" y=\"-528.8\">x_2 ≤ -0.025</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2697.5\" y=\"-513.8\">entropy = 0.999</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2705\" y=\"-498.8\">samples = 23</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2690.5\" y=\"-483.8\">value = [0, 11, 12]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2707.5\" y=\"-468.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 94&#45;&gt;102 -->\n",
       "<g class=\"edge\" id=\"edge102\">\n",
       "<title>94-&gt;102</title>\n",
       "<path d=\"M2655.3667,-579.8796C2665.8125,-570.6031 2676.9781,-560.6874 2687.711,-551.1559\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2690.2801,-553.5553 2695.4333,-544.2981 2685.632,-548.3213 2690.2801,-553.5553\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 96 -->\n",
       "<g class=\"node\" id=\"node97\">\n",
       "<title>96</title>\n",
       "<polygon fill=\"#d0f9e1\" points=\"2553.5,-425 2433.5,-425 2433.5,-342 2553.5,-342 2553.5,-425\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2459.5\" y=\"-409.8\">x_1 ≤ 0.204</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2448.5\" y=\"-394.8\">entropy = 0.987</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2456\" y=\"-379.8\">samples = 30</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2441.5\" y=\"-364.8\">value = [0, 17, 13]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2456.5\" y=\"-349.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 95&#45;&gt;96 -->\n",
       "<g class=\"edge\" id=\"edge96\">\n",
       "<title>95-&gt;96</title>\n",
       "<path d=\"M2565.6776,-460.8796C2557.1927,-451.7832 2548.1346,-442.0722 2539.4032,-432.7116\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2541.8686,-430.2233 2532.4881,-425.2981 2536.7497,-434.9981 2541.8686,-430.2233\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 101 -->\n",
       "<g class=\"node\" id=\"node102\">\n",
       "<title>101</title>\n",
       "<polygon fill=\"#39e581\" points=\"2685,-417.5 2572,-417.5 2572,-349.5 2685,-349.5 2685,-417.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2590.5\" y=\"-402.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2591\" y=\"-387.3\">samples = 10</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2580\" y=\"-372.3\">value = [0, 10, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2591.5\" y=\"-357.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 95&#45;&gt;101 -->\n",
       "<g class=\"edge\" id=\"edge101\">\n",
       "<title>95-&gt;101</title>\n",
       "<path d=\"M2612.894,-460.8796C2615.0452,-450.2134 2617.3668,-438.7021 2619.5451,-427.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2623.0332,-428.3095 2621.5793,-417.8149 2616.1713,-426.9256 2623.0332,-428.3095\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 97 -->\n",
       "<g class=\"node\" id=\"node98\">\n",
       "<title>97</title>\n",
       "<polygon fill=\"#8bf0b5\" points=\"2729,-306 2616,-306 2616,-223 2729,-223 2729,-306\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2638.5\" y=\"-290.8\">x_1 ≤ 0.182</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2627.5\" y=\"-275.8\">entropy = 0.874</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2635\" y=\"-260.8\">samples = 17</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2624\" y=\"-245.8\">value = [0, 12, 5]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2635.5\" y=\"-230.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 96&#45;&gt;97 -->\n",
       "<g class=\"edge\" id=\"edge97\">\n",
       "<title>96-&gt;97</title>\n",
       "<path d=\"M2553.6024,-343.5437C2570.8843,-332.0546 2589.8114,-319.4718 2607.3813,-307.7912\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2609.4944,-310.5893 2615.8844,-302.1383 2605.619,-304.7599 2609.4944,-310.5893\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 100 -->\n",
       "<g class=\"node\" id=\"node101\">\n",
       "<title>100</title>\n",
       "<polygon fill=\"#d0b5f5\" points=\"2853.5,-298.5 2747.5,-298.5 2747.5,-230.5 2853.5,-230.5 2853.5,-298.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2755.5\" y=\"-283.3\">entropy = 0.961</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2763\" y=\"-268.3\">samples = 13</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2755.5\" y=\"-253.3\">value = [0, 5, 8]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2765.5\" y=\"-238.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 96&#45;&gt;100 -->\n",
       "<g class=\"edge\" id=\"edge100\">\n",
       "<title>96-&gt;100</title>\n",
       "<path d=\"M2553.8781,-345.67C2556.7532,-344.351 2559.6352,-343.1184 2562.5,-342 2636.4699,-313.1238 2664.1376,-336.3866 2737.5,-306 2739.5646,-305.1448 2741.6339,-304.2193 2743.698,-303.2366\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2745.4011,-306.2964 2752.6849,-298.6025 2742.1929,-300.0749 2745.4011,-306.2964\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 98 -->\n",
       "<g class=\"node\" id=\"node99\">\n",
       "<title>98</title>\n",
       "<polygon fill=\"#72eca5\" points=\"2754.5,-179.5 2648.5,-179.5 2648.5,-111.5 2754.5,-111.5 2754.5,-179.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2656.5\" y=\"-164.3\">entropy = 0.764</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2667.5\" y=\"-149.3\">samples = 9</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2656.5\" y=\"-134.3\">value = [0, 7, 2]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2664.5\" y=\"-119.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 97&#45;&gt;98 -->\n",
       "<g class=\"edge\" id=\"edge98\">\n",
       "<title>97-&gt;98</title>\n",
       "<path d=\"M2682.6428,-222.8796C2685.2689,-212.1034 2688.1052,-200.4647 2690.7608,-189.5677\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2694.1703,-190.3593 2693.1375,-179.8149 2687.3693,-188.7019 2694.1703,-190.3593\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 99 -->\n",
       "<g class=\"node\" id=\"node100\">\n",
       "<title>99</title>\n",
       "<polygon fill=\"#b0f5cd\" points=\"2878.5,-179.5 2772.5,-179.5 2772.5,-111.5 2878.5,-111.5 2878.5,-179.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2780.5\" y=\"-164.3\">entropy = 0.954</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2791.5\" y=\"-149.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2780.5\" y=\"-134.3\">value = [0, 5, 3]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2788.5\" y=\"-119.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 97&#45;&gt;99 -->\n",
       "<g class=\"edge\" id=\"edge99\">\n",
       "<title>97-&gt;99</title>\n",
       "<path d=\"M2726.012,-222.8796C2741.2808,-211.0038 2757.8966,-198.0804 2773.0984,-186.2568\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2775.6361,-188.7171 2781.3808,-179.8149 2771.3384,-183.1916 2775.6361,-188.7171\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 103 -->\n",
       "<g class=\"node\" id=\"node104\">\n",
       "<title>103</title>\n",
       "<polygon fill=\"#b0f5cd\" points=\"2809.5,-417.5 2703.5,-417.5 2703.5,-349.5 2809.5,-349.5 2809.5,-417.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2711.5\" y=\"-402.3\">entropy = 0.954</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2722.5\" y=\"-387.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2711.5\" y=\"-372.3\">value = [0, 5, 3]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2719.5\" y=\"-357.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 102&#45;&gt;103 -->\n",
       "<g class=\"edge\" id=\"edge103\">\n",
       "<title>102-&gt;103</title>\n",
       "<path d=\"M2747.3965,-460.8796C2748.6514,-450.2134 2750.0056,-438.7021 2751.2763,-427.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2754.7705,-428.1554 2752.4629,-417.8149 2747.8184,-427.3375 2754.7705,-428.1554\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 104 -->\n",
       "<g class=\"node\" id=\"node105\">\n",
       "<title>104</title>\n",
       "<polygon fill=\"#d5bdf6\" points=\"2933.5,-417.5 2827.5,-417.5 2827.5,-349.5 2933.5,-349.5 2933.5,-417.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2835.5\" y=\"-402.3\">entropy = 0.971</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2843\" y=\"-387.3\">samples = 15</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2835.5\" y=\"-372.3\">value = [0, 6, 9]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2845.5\" y=\"-357.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 102&#45;&gt;104 -->\n",
       "<g class=\"edge\" id=\"edge104\">\n",
       "<title>102-&gt;104</title>\n",
       "<path d=\"M2790.7657,-460.8796C2804.4101,-449.1138 2819.2469,-436.3197 2832.8546,-424.5855\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2835.4187,-426.9961 2840.7062,-417.8149 2830.8473,-421.6948 2835.4187,-426.9961\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 106 -->\n",
       "<g class=\"node\" id=\"node107\">\n",
       "<title>106</title>\n",
       "<polygon fill=\"#8ef0b7\" points=\"2800,-663 2687,-663 2687,-580 2800,-580 2800,-663\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2707.5\" y=\"-647.8\">x_2 ≤ -0.209</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2698.5\" y=\"-632.8\">entropy = 0.881</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2706\" y=\"-617.8\">samples = 20</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2695\" y=\"-602.8\">value = [0, 14, 6]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2706.5\" y=\"-587.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 105&#45;&gt;106 -->\n",
       "<g class=\"edge\" id=\"edge106\">\n",
       "<title>105-&gt;106</title>\n",
       "<path d=\"M2743.5,-698.8796C2743.5,-690.6838 2743.5,-681.9891 2743.5,-673.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2747.0001,-673.298 2743.5,-663.2981 2740.0001,-673.2981 2747.0001,-673.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 109 -->\n",
       "<g class=\"node\" id=\"node110\">\n",
       "<title>109</title>\n",
       "<polygon fill=\"#39e581\" points=\"2931,-655.5 2818,-655.5 2818,-587.5 2931,-587.5 2931,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2836.5\" y=\"-640.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2837\" y=\"-625.3\">samples = 17</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2826\" y=\"-610.3\">value = [0, 17, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2837.5\" y=\"-595.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 105&#45;&gt;109 -->\n",
       "<g class=\"edge\" id=\"edge109\">\n",
       "<title>105-&gt;109</title>\n",
       "<path d=\"M2789.3174,-698.8796C2802.2697,-687.1138 2816.354,-674.3197 2829.2714,-662.5855\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2831.6761,-665.1296 2836.7247,-655.8149 2826.9694,-659.9482 2831.6761,-665.1296\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 107 -->\n",
       "<g class=\"node\" id=\"node108\">\n",
       "<title>107</title>\n",
       "<polygon fill=\"#39e581\" points=\"2934,-536.5 2821,-536.5 2821,-468.5 2934,-468.5 2934,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2839.5\" y=\"-521.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2840\" y=\"-506.3\">samples = 11</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2829\" y=\"-491.3\">value = [0, 11, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2840.5\" y=\"-476.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 106&#45;&gt;107 -->\n",
       "<g class=\"edge\" id=\"edge107\">\n",
       "<title>106-&gt;107</title>\n",
       "<path d=\"M2790.3667,-579.8796C2803.6156,-568.1138 2818.0224,-555.3197 2831.2356,-543.5855\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2833.7065,-546.0722 2838.8597,-536.8149 2829.0584,-540.8381 2833.7065,-546.0722\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 108 -->\n",
       "<g class=\"node\" id=\"node109\">\n",
       "<title>108</title>\n",
       "<polygon fill=\"#c09cf2\" points=\"3058.5,-536.5 2952.5,-536.5 2952.5,-468.5 3058.5,-468.5 3058.5,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2960.5\" y=\"-521.3\">entropy = 0.918</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2971.5\" y=\"-506.3\">samples = 9</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2960.5\" y=\"-491.3\">value = [0, 3, 6]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2970.5\" y=\"-476.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 106&#45;&gt;108 -->\n",
       "<g class=\"edge\" id=\"edge108\">\n",
       "<title>106-&gt;108</title>\n",
       "<path d=\"M2800.0182,-584.0326C2802.8465,-582.5968 2805.6824,-581.2425 2808.5,-580 2864.9244,-555.1171 2886.2546,-569.2847 2942.5,-544 2944.2919,-543.1945 2946.09,-542.3409 2947.8876,-541.4476\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"2949.6517,-544.4737 2956.8388,-536.6894 2946.366,-538.2927 2949.6517,-544.4737\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 111 -->\n",
       "<g class=\"node\" id=\"node112\">\n",
       "<title>111</title>\n",
       "<polygon fill=\"#eca26d\" points=\"3255.5,-1020 3135.5,-1020 3135.5,-937 3255.5,-937 3255.5,-1020\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3159.5\" y=\"-1004.8\">x_2 ≤ -0.566</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3150.5\" y=\"-989.8\">entropy = 0.735</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3158\" y=\"-974.8\">samples = 92</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3143.5\" y=\"-959.8\">value = [73, 19, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3165.5\" y=\"-944.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 110&#45;&gt;111 -->\n",
       "<g class=\"edge\" id=\"edge111\">\n",
       "<title>110-&gt;111</title>\n",
       "<path d=\"M3195.5,-1055.8796C3195.5,-1047.6838 3195.5,-1038.9891 3195.5,-1030.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3199.0001,-1030.298 3195.5,-1020.2981 3192.0001,-1030.2981 3199.0001,-1030.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 120 -->\n",
       "<g class=\"node\" id=\"node121\">\n",
       "<title>120</title>\n",
       "<polygon fill=\"#e5fcee\" points=\"3642,-1020 3515,-1020 3515,-937 3642,-937 3642,-1020\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3544.5\" y=\"-1004.8\">x_2 ≤ 0.309</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3533.5\" y=\"-989.8\">entropy = 1.565</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3537.5\" y=\"-974.8\">samples = 213</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3523\" y=\"-959.8\">value = [68, 87, 58]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3541.5\" y=\"-944.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 110&#45;&gt;120 -->\n",
       "<g class=\"edge\" id=\"edge120\">\n",
       "<title>110-&gt;120</title>\n",
       "<path d=\"M3266.1944,-1075.5349C3334.0754,-1054.4439 3436.185,-1022.718 3504.9925,-1001.3391\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3506.3836,-1004.572 3514.8948,-998.2625 3504.3066,-997.8873 3506.3836,-1004.572\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 112 -->\n",
       "<g class=\"node\" id=\"node113\">\n",
       "<title>112</title>\n",
       "<polygon fill=\"#e89050\" points=\"3184,-901 3071,-901 3071,-818 3184,-818 3184,-901\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3093.5\" y=\"-885.8\">x_1 ≤ 0.561</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3082.5\" y=\"-870.8\">entropy = 0.483</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3090\" y=\"-855.8\">samples = 67</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3079\" y=\"-840.8\">value = [60, 7, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3097.5\" y=\"-825.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 111&#45;&gt;112 -->\n",
       "<g class=\"edge\" id=\"edge112\">\n",
       "<title>111-&gt;112</title>\n",
       "<path d=\"M3171.7169,-936.8796C3166.7763,-928.2335 3161.5184,-919.0322 3156.4167,-910.1042\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3159.3849,-908.244 3151.3846,-901.2981 3153.3072,-911.717 3159.3849,-908.244\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 117 -->\n",
       "<g class=\"node\" id=\"node118\">\n",
       "<title>117</title>\n",
       "<polygon fill=\"#fdf5f0\" points=\"3322.5,-901 3202.5,-901 3202.5,-818 3322.5,-818 3322.5,-901\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3226.5\" y=\"-885.8\">x_2 ≤ -0.496</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3217.5\" y=\"-870.8\">entropy = 0.999</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3225\" y=\"-855.8\">samples = 25</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3210.5\" y=\"-840.8\">value = [13, 12, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3232.5\" y=\"-825.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 111&#45;&gt;117 -->\n",
       "<g class=\"edge\" id=\"edge117\">\n",
       "<title>111-&gt;117</title>\n",
       "<path d=\"M3218.9333,-936.8796C3223.8013,-928.2335 3228.9819,-919.0322 3234.0086,-910.1042\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3237.1103,-911.729 3238.9666,-901.2981 3231.0107,-908.2947 3237.1103,-911.729\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 113 -->\n",
       "<g class=\"node\" id=\"node114\">\n",
       "<title>113</title>\n",
       "<polygon fill=\"#f9e3d3\" points=\"3055.5,-782 2949.5,-782 2949.5,-699 3055.5,-699 3055.5,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2970\" y=\"-766.8\">x_2 ≤ -0.81</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2957.5\" y=\"-751.8\">entropy = 0.989</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2965\" y=\"-736.8\">samples = 16</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2957.5\" y=\"-721.8\">value = [9, 7, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2972.5\" y=\"-706.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 112&#45;&gt;113 -->\n",
       "<g class=\"edge\" id=\"edge113\">\n",
       "<title>112-&gt;113</title>\n",
       "<path d=\"M3083.7811,-817.8796C3074.1314,-808.6931 3063.8234,-798.8798 3053.9009,-789.4336\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3056.0616,-786.6582 3046.4055,-782.2981 3051.235,-791.7282 3056.0616,-786.6582\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 116 -->\n",
       "<g class=\"node\" id=\"node117\">\n",
       "<title>116</title>\n",
       "<polygon fill=\"#e58139\" points=\"3187,-774.5 3074,-774.5 3074,-706.5 3187,-706.5 3187,-774.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3092.5\" y=\"-759.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3093\" y=\"-744.3\">samples = 51</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3082\" y=\"-729.3\">value = [51, 0, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3100.5\" y=\"-714.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 112&#45;&gt;116 -->\n",
       "<g class=\"edge\" id=\"edge116\">\n",
       "<title>112-&gt;116</title>\n",
       "<path d=\"M3128.5493,-817.8796C3128.8182,-807.2134 3129.1084,-795.7021 3129.3806,-784.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3132.8817,-784.9 3129.6349,-774.8149 3125.8839,-784.7235 3132.8817,-784.9\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 114 -->\n",
       "<g class=\"node\" id=\"node115\">\n",
       "<title>114</title>\n",
       "<polygon fill=\"#eeab7b\" points=\"3055.5,-655.5 2949.5,-655.5 2949.5,-587.5 3055.5,-587.5 3055.5,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2957.5\" y=\"-640.3\">entropy = 0.811</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2968.5\" y=\"-625.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2957.5\" y=\"-610.3\">value = [6, 2, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"2972.5\" y=\"-595.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 113&#45;&gt;114 -->\n",
       "<g class=\"edge\" id=\"edge114\">\n",
       "<title>113-&gt;114</title>\n",
       "<path d=\"M3002.5,-698.8796C3002.5,-688.2134 3002.5,-676.7021 3002.5,-665.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3006.0001,-665.8149 3002.5,-655.8149 2999.0001,-665.815 3006.0001,-665.8149\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 115 -->\n",
       "<g class=\"node\" id=\"node116\">\n",
       "<title>115</title>\n",
       "<polygon fill=\"#b0f5cd\" points=\"3179.5,-655.5 3073.5,-655.5 3073.5,-587.5 3179.5,-587.5 3179.5,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3081.5\" y=\"-640.3\">entropy = 0.954</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3092.5\" y=\"-625.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3081.5\" y=\"-610.3\">value = [3, 5, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3089.5\" y=\"-595.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 113&#45;&gt;115 -->\n",
       "<g class=\"edge\" id=\"edge115\">\n",
       "<title>113-&gt;115</title>\n",
       "<path d=\"M3045.8692,-698.8796C3058.0148,-687.2237 3071.2119,-674.5587 3083.3451,-662.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3085.9516,-665.2643 3090.7433,-655.8149 3081.1047,-660.2138 3085.9516,-665.2643\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 118 -->\n",
       "<g class=\"node\" id=\"node119\">\n",
       "<title>118</title>\n",
       "<polygon fill=\"#aaf4c9\" points=\"3311.5,-774.5 3205.5,-774.5 3205.5,-706.5 3311.5,-706.5 3311.5,-774.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3213.5\" y=\"-759.3\">entropy = 0.946</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3221\" y=\"-744.3\">samples = 11</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3213.5\" y=\"-729.3\">value = [4, 7, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3221.5\" y=\"-714.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 117&#45;&gt;118 -->\n",
       "<g class=\"edge\" id=\"edge118\">\n",
       "<title>117-&gt;118</title>\n",
       "<path d=\"M3261.101,-817.8796C3260.7425,-807.2134 3260.3555,-795.7021 3259.9925,-784.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3263.4875,-784.6917 3259.6534,-774.8149 3256.4915,-784.9269 3263.4875,-784.6917\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 119 -->\n",
       "<g class=\"node\" id=\"node120\">\n",
       "<title>119</title>\n",
       "<polygon fill=\"#f3c7a7\" points=\"3435.5,-774.5 3329.5,-774.5 3329.5,-706.5 3435.5,-706.5 3435.5,-774.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3341\" y=\"-759.3\">entropy = 0.94</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3345\" y=\"-744.3\">samples = 14</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3337.5\" y=\"-729.3\">value = [9, 5, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3352.5\" y=\"-714.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 117&#45;&gt;119 -->\n",
       "<g class=\"edge\" id=\"edge119\">\n",
       "<title>117-&gt;119</title>\n",
       "<path d=\"M3304.4702,-817.8796C3316.224,-806.2237 3328.9954,-793.5587 3340.7372,-781.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3343.2606,-784.3416 3347.8967,-774.8149 3338.3316,-779.3712 3343.2606,-784.3416\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 121 -->\n",
       "<g class=\"node\" id=\"node122\">\n",
       "<title>121</title>\n",
       "<polygon fill=\"#ede3fb\" points=\"3638.5,-901 3518.5,-901 3518.5,-818 3638.5,-818 3638.5,-901\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3542.5\" y=\"-885.8\">x_2 ≤ -0.182</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3533.5\" y=\"-870.8\">entropy = 1.214</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3537.5\" y=\"-855.8\">samples = 112</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3526.5\" y=\"-840.8\">value = [5, 49, 58]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3543.5\" y=\"-825.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 120&#45;&gt;121 -->\n",
       "<g class=\"edge\" id=\"edge121\">\n",
       "<title>120-&gt;121</title>\n",
       "<path d=\"M3578.5,-936.8796C3578.5,-928.6838 3578.5,-919.9891 3578.5,-911.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3582.0001,-911.298 3578.5,-901.2981 3575.0001,-911.2981 3582.0001,-911.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 142 -->\n",
       "<g class=\"node\" id=\"node143\">\n",
       "<title>142</title>\n",
       "<polygon fill=\"#f5cdb0\" points=\"4027.5,-901 3907.5,-901 3907.5,-818 4027.5,-818 4027.5,-901\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3933.5\" y=\"-885.8\">x_2 ≤ 0.462</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3922.5\" y=\"-870.8\">entropy = 0.955</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3926.5\" y=\"-855.8\">samples = 101</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3915.5\" y=\"-840.8\">value = [63, 38, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3937.5\" y=\"-825.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 120&#45;&gt;142 -->\n",
       "<g class=\"edge\" id=\"edge142\">\n",
       "<title>120-&gt;142</title>\n",
       "<path d=\"M3642.2998,-958.9828C3712.3955,-937.5397 3824.6433,-903.2017 3897.2173,-881.0004\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3898.6472,-884.2231 3907.1859,-877.9508 3896.5995,-877.5293 3898.6472,-884.2231\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 122 -->\n",
       "<g class=\"node\" id=\"node123\">\n",
       "<title>122</title>\n",
       "<polygon fill=\"#74eda7\" points=\"3567,-782 3454,-782 3454,-699 3567,-699 3567,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3474.5\" y=\"-766.8\">x_2 ≤ -0.237</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3465.5\" y=\"-751.8\">entropy = 1.041</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3473\" y=\"-736.8\">samples = 36</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3462\" y=\"-721.8\">value = [3, 27, 6]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3473.5\" y=\"-706.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 121&#45;&gt;122 -->\n",
       "<g class=\"edge\" id=\"edge122\">\n",
       "<title>121-&gt;122</title>\n",
       "<path d=\"M3554.7169,-817.8796C3549.7763,-809.2335 3544.5184,-800.0322 3539.4167,-791.1042\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3542.3849,-789.244 3534.3846,-782.2981 3536.3072,-792.717 3542.3849,-789.244\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 129 -->\n",
       "<g class=\"node\" id=\"node130\">\n",
       "<title>129</title>\n",
       "<polygon fill=\"#b991f1\" points=\"3705.5,-782 3585.5,-782 3585.5,-699 3705.5,-699 3705.5,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3611.5\" y=\"-766.8\">x_2 ≤ 0.218</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3604\" y=\"-751.8\">entropy = 1.03</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3608\" y=\"-736.8\">samples = 76</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3593.5\" y=\"-721.8\">value = [2, 22, 52]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3610.5\" y=\"-706.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 121&#45;&gt;129 -->\n",
       "<g class=\"edge\" id=\"edge129\">\n",
       "<title>121-&gt;129</title>\n",
       "<path d=\"M3601.9333,-817.8796C3606.8013,-809.2335 3611.9819,-800.0322 3617.0086,-791.1042\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3620.1103,-792.729 3621.9666,-782.2981 3614.0107,-789.2947 3620.1103,-792.729\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 123 -->\n",
       "<g class=\"node\" id=\"node124\">\n",
       "<title>123</title>\n",
       "<polygon fill=\"#6deca2\" points=\"3439,-663 3326,-663 3326,-580 3439,-580 3439,-663\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3348.5\" y=\"-647.8\">x_1 ≤ 0.549</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3337.5\" y=\"-632.8\">entropy = 0.889</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3345\" y=\"-617.8\">samples = 28</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3334\" y=\"-602.8\">value = [1, 22, 5]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3345.5\" y=\"-587.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 122&#45;&gt;123 -->\n",
       "<g class=\"edge\" id=\"edge123\">\n",
       "<title>122-&gt;123</title>\n",
       "<path d=\"M3465.7318,-698.8796C3455.7537,-689.6031 3445.0881,-679.6874 3434.8358,-670.1559\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3437.1663,-667.5436 3427.4593,-663.2981 3432.4,-672.6704 3437.1663,-667.5436\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 128 -->\n",
       "<g class=\"node\" id=\"node129\">\n",
       "<title>128</title>\n",
       "<polygon fill=\"#9cf2c0\" points=\"3563.5,-655.5 3457.5,-655.5 3457.5,-587.5 3563.5,-587.5 3563.5,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3465.5\" y=\"-640.3\">entropy = 1.299</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3476.5\" y=\"-625.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3465.5\" y=\"-610.3\">value = [2, 5, 1]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3473.5\" y=\"-595.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 122&#45;&gt;128 -->\n",
       "<g class=\"edge\" id=\"edge128\">\n",
       "<title>122-&gt;128</title>\n",
       "<path d=\"M3510.5,-698.8796C3510.5,-688.2134 3510.5,-676.7021 3510.5,-665.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3514.0001,-665.8149 3510.5,-655.8149 3507.0001,-665.815 3514.0001,-665.8149\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 124 -->\n",
       "<g class=\"node\" id=\"node125\">\n",
       "<title>124</title>\n",
       "<polygon fill=\"#39e581\" points=\"3307.5,-536.5 3201.5,-536.5 3201.5,-468.5 3307.5,-468.5 3307.5,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3216.5\" y=\"-521.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3220.5\" y=\"-506.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3209.5\" y=\"-491.3\">value = [0, 8, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3217.5\" y=\"-476.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 123&#45;&gt;124 -->\n",
       "<g class=\"edge\" id=\"edge124\">\n",
       "<title>123-&gt;124</title>\n",
       "<path d=\"M3337.7318,-579.8796C3325.1944,-568.2237 3311.5716,-555.5587 3299.047,-543.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3301.1172,-541.0605 3291.4102,-536.8149 3296.3509,-546.1872 3301.1172,-541.0605\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 125 -->\n",
       "<g class=\"node\" id=\"node126\">\n",
       "<title>125</title>\n",
       "<polygon fill=\"#88efb3\" points=\"3439,-544 3326,-544 3326,-461 3439,-461 3439,-544\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3346.5\" y=\"-528.8\">x_2 ≤ -0.298</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3337.5\" y=\"-513.8\">entropy = 1.076</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3345\" y=\"-498.8\">samples = 20</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3334\" y=\"-483.8\">value = [1, 14, 5]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3345.5\" y=\"-468.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 123&#45;&gt;125 -->\n",
       "<g class=\"edge\" id=\"edge125\">\n",
       "<title>123-&gt;125</title>\n",
       "<path d=\"M3382.5,-579.8796C3382.5,-571.6838 3382.5,-562.9891 3382.5,-554.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3386.0001,-554.298 3382.5,-544.2981 3379.0001,-554.2981 3386.0001,-554.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 126 -->\n",
       "<g class=\"node\" id=\"node127\">\n",
       "<title>126</title>\n",
       "<polygon fill=\"#65eb9d\" points=\"3312.5,-417.5 3206.5,-417.5 3206.5,-349.5 3312.5,-349.5 3312.5,-417.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3214.5\" y=\"-402.3\">entropy = 0.922</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3222\" y=\"-387.3\">samples = 10</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3214.5\" y=\"-372.3\">value = [1, 8, 1]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3222.5\" y=\"-357.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 125&#45;&gt;126 -->\n",
       "<g class=\"edge\" id=\"edge126\">\n",
       "<title>125-&gt;126</title>\n",
       "<path d=\"M3339.4806,-460.8796C3327.4329,-449.2237 3314.3422,-436.5587 3302.3069,-424.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3304.589,-422.2528 3294.9684,-417.8149 3299.7217,-427.2837 3304.589,-422.2528\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 127 -->\n",
       "<g class=\"node\" id=\"node128\">\n",
       "<title>127</title>\n",
       "<polygon fill=\"#bdf6d5\" points=\"3436.5,-417.5 3330.5,-417.5 3330.5,-349.5 3436.5,-349.5 3436.5,-417.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3338.5\" y=\"-402.3\">entropy = 0.971</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3346\" y=\"-387.3\">samples = 10</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3338.5\" y=\"-372.3\">value = [0, 6, 4]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3346.5\" y=\"-357.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 125&#45;&gt;127 -->\n",
       "<g class=\"edge\" id=\"edge127\">\n",
       "<title>125-&gt;127</title>\n",
       "<path d=\"M3382.8498,-460.8796C3382.9394,-450.2134 3383.0361,-438.7021 3383.1269,-427.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3386.6274,-427.844 3383.2116,-417.8149 3379.6276,-427.7851 3386.6274,-427.844\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 130 -->\n",
       "<g class=\"node\" id=\"node131\">\n",
       "<title>130</title>\n",
       "<polygon fill=\"#9e67eb\" points=\"3701.5,-663 3581.5,-663 3581.5,-580 3701.5,-580 3701.5,-663\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3607.5\" y=\"-647.8\">x_2 ≤ 0.105</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3596.5\" y=\"-632.8\">entropy = 0.781</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3604\" y=\"-617.8\">samples = 63</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3589.5\" y=\"-602.8\">value = [1, 11, 51]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3606.5\" y=\"-587.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 129&#45;&gt;130 -->\n",
       "<g class=\"edge\" id=\"edge130\">\n",
       "<title>129-&gt;130</title>\n",
       "<path d=\"M3644.101,-698.8796C3643.8255,-690.6838 3643.5332,-681.9891 3643.2479,-673.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3646.739,-673.1748 3642.905,-663.2981 3639.743,-673.41 3646.739,-673.1748\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 141 -->\n",
       "<g class=\"node\" id=\"node142\">\n",
       "<title>141</title>\n",
       "<polygon fill=\"#5ae996\" points=\"3833,-655.5 3720,-655.5 3720,-587.5 3833,-587.5 3833,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3731.5\" y=\"-640.3\">entropy = 0.773</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3739\" y=\"-625.3\">samples = 13</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3728\" y=\"-610.3\">value = [1, 11, 1]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3739.5\" y=\"-595.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 129&#45;&gt;141 -->\n",
       "<g class=\"edge\" id=\"edge141\">\n",
       "<title>129-&gt;141</title>\n",
       "<path d=\"M3691.3174,-698.8796C3704.2697,-687.1138 3718.354,-674.3197 3731.2714,-662.5855\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3733.6761,-665.1296 3738.7247,-655.8149 3728.9694,-659.9482 3733.6761,-665.1296\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 131 -->\n",
       "<g class=\"node\" id=\"node132\">\n",
       "<title>131</title>\n",
       "<polygon fill=\"#9051e8\" points=\"3698,-544 3585,-544 3585,-461 3698,-461 3698,-544\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3607.5\" y=\"-528.8\">x_2 ≤ 0.013</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3596.5\" y=\"-513.8\">entropy = 0.574</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3604\" y=\"-498.8\">samples = 46</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3593\" y=\"-483.8\">value = [1, 4, 41]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3606.5\" y=\"-468.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 130&#45;&gt;131 -->\n",
       "<g class=\"edge\" id=\"edge131\">\n",
       "<title>130-&gt;131</title>\n",
       "<path d=\"M3641.5,-579.8796C3641.5,-571.6838 3641.5,-562.9891 3641.5,-554.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3645.0001,-554.298 3641.5,-544.2981 3638.0001,-554.2981 3645.0001,-554.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 138 -->\n",
       "<g class=\"node\" id=\"node139\">\n",
       "<title>138</title>\n",
       "<polygon fill=\"#d9c4f7\" points=\"3829,-544 3716,-544 3716,-461 3829,-461 3829,-544\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3742\" y=\"-528.8\">x_1 ≤ 0.75</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3727.5\" y=\"-513.8\">entropy = 0.977</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3735\" y=\"-498.8\">samples = 17</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3724\" y=\"-483.8\">value = [0, 7, 10]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3737.5\" y=\"-468.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 130&#45;&gt;138 -->\n",
       "<g class=\"edge\" id=\"edge138\">\n",
       "<title>130-&gt;138</title>\n",
       "<path d=\"M3687.3174,-579.8796C3697.5294,-570.6031 3708.445,-560.6874 3718.9376,-551.1559\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3721.4384,-553.6127 3726.487,-544.2981 3716.7316,-548.4313 3721.4384,-553.6127\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 132 -->\n",
       "<g class=\"node\" id=\"node133\">\n",
       "<title>132</title>\n",
       "<polygon fill=\"#985eea\" points=\"3569,-425 3456,-425 3456,-342 3569,-342 3569,-425\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3478.5\" y=\"-409.8\">x_1 ≤ 0.755</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3467.5\" y=\"-394.8\">entropy = 0.754</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3475\" y=\"-379.8\">samples = 31</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3464\" y=\"-364.8\">value = [1, 4, 26]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3477.5\" y=\"-349.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 131&#45;&gt;132 -->\n",
       "<g class=\"edge\" id=\"edge132\">\n",
       "<title>131-&gt;132</title>\n",
       "<path d=\"M3596.3821,-460.8796C3586.326,-451.6031 3575.577,-441.6874 3565.2446,-432.1559\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3567.5339,-429.506 3557.8105,-425.2981 3562.7876,-434.6511 3567.5339,-429.506\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 137 -->\n",
       "<g class=\"node\" id=\"node138\">\n",
       "<title>137</title>\n",
       "<polygon fill=\"#8139e5\" points=\"3700,-417.5 3587,-417.5 3587,-349.5 3700,-349.5 3700,-417.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3605.5\" y=\"-402.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3606\" y=\"-387.3\">samples = 15</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3595\" y=\"-372.3\">value = [0, 0, 15]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3608.5\" y=\"-357.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 131&#45;&gt;137 -->\n",
       "<g class=\"edge\" id=\"edge137\">\n",
       "<title>131-&gt;137</title>\n",
       "<path d=\"M3642.1995,-460.8796C3642.3788,-450.2134 3642.5722,-438.7021 3642.7538,-427.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3646.2546,-427.8724 3642.9233,-417.8149 3639.2556,-427.7547 3646.2546,-427.8724\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 133 -->\n",
       "<g class=\"node\" id=\"node134\">\n",
       "<title>133</title>\n",
       "<polygon fill=\"#a470ec\" points=\"3505,-306 3392,-306 3392,-223 3505,-223 3505,-306\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3414.5\" y=\"-290.8\">x_1 ≤ 0.464</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3403.5\" y=\"-275.8\">entropy = 0.937</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3411\" y=\"-260.8\">samples = 22</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3400\" y=\"-245.8\">value = [1, 4, 17]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3413.5\" y=\"-230.8\">class = navy</text>\n",
       "</g>\n",
       "<!-- 132&#45;&gt;133 -->\n",
       "<g class=\"edge\" id=\"edge133\">\n",
       "<title>132-&gt;133</title>\n",
       "<path d=\"M3490.1159,-341.8796C3485.5144,-333.3236 3480.6204,-324.2238 3475.8658,-315.3833\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3478.7988,-313.4473 3470.9796,-306.2981 3472.6338,-316.763 3478.7988,-313.4473\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 136 -->\n",
       "<g class=\"node\" id=\"node137\">\n",
       "<title>136</title>\n",
       "<polygon fill=\"#8139e5\" points=\"3629.5,-298.5 3523.5,-298.5 3523.5,-230.5 3629.5,-230.5 3629.5,-298.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3538.5\" y=\"-283.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3542.5\" y=\"-268.3\">samples = 9</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3531.5\" y=\"-253.3\">value = [0, 0, 9]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3541.5\" y=\"-238.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 132&#45;&gt;136 -->\n",
       "<g class=\"edge\" id=\"edge136\">\n",
       "<title>132-&gt;136</title>\n",
       "<path d=\"M3534.8841,-341.8796C3540.798,-330.8835 3547.1949,-318.9893 3553.158,-307.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3556.3908,-309.2799 3558.0449,-298.8149 3550.2258,-305.9642 3556.3908,-309.2799\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 134 -->\n",
       "<g class=\"node\" id=\"node135\">\n",
       "<title>134</title>\n",
       "<polygon fill=\"#8139e5\" points=\"3439.5,-179.5 3333.5,-179.5 3333.5,-111.5 3439.5,-111.5 3439.5,-179.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3348.5\" y=\"-164.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3352.5\" y=\"-149.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3341.5\" y=\"-134.3\">value = [0, 0, 8]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3351.5\" y=\"-119.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 133&#45;&gt;134 -->\n",
       "<g class=\"edge\" id=\"edge134\">\n",
       "<title>133-&gt;134</title>\n",
       "<path d=\"M3426.8154,-222.8796C3421.0864,-211.8835 3414.8894,-199.9893 3409.1126,-188.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3412.103,-187.0662 3404.3784,-179.8149 3405.895,-190.3007 3412.103,-187.0662\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 135 -->\n",
       "<g class=\"node\" id=\"node136\">\n",
       "<title>135</title>\n",
       "<polygon fill=\"#c09cf2\" points=\"3563.5,-179.5 3457.5,-179.5 3457.5,-111.5 3563.5,-111.5 3563.5,-179.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3465.5\" y=\"-164.3\">entropy = 1.198</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3473\" y=\"-149.3\">samples = 14</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3465.5\" y=\"-134.3\">value = [1, 4, 9]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3475.5\" y=\"-119.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 133&#45;&gt;135 -->\n",
       "<g class=\"edge\" id=\"edge135\">\n",
       "<title>133-&gt;135</title>\n",
       "<path d=\"M3470.1846,-222.8796C3475.9136,-211.8835 3482.1106,-199.9893 3487.8874,-188.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3491.105,-190.3007 3492.6216,-179.8149 3484.897,-187.0662 3491.105,-190.3007\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 139 -->\n",
       "<g class=\"node\" id=\"node140\">\n",
       "<title>139</title>\n",
       "<polygon fill=\"#c09cf2\" points=\"3825.5,-417.5 3719.5,-417.5 3719.5,-349.5 3825.5,-349.5 3825.5,-417.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3727.5\" y=\"-402.3\">entropy = 0.918</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3738.5\" y=\"-387.3\">samples = 9</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3727.5\" y=\"-372.3\">value = [0, 3, 6]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3737.5\" y=\"-357.3\">class = navy</text>\n",
       "</g>\n",
       "<!-- 138&#45;&gt;139 -->\n",
       "<g class=\"edge\" id=\"edge139\">\n",
       "<title>138-&gt;139</title>\n",
       "<path d=\"M3772.5,-460.8796C3772.5,-450.2134 3772.5,-438.7021 3772.5,-427.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3776.0001,-427.8149 3772.5,-417.8149 3769.0001,-427.815 3776.0001,-427.8149\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 140 -->\n",
       "<g class=\"node\" id=\"node141\">\n",
       "<title>140</title>\n",
       "<polygon fill=\"#ffffff\" points=\"3949.5,-417.5 3843.5,-417.5 3843.5,-349.5 3949.5,-349.5 3949.5,-417.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3858.5\" y=\"-402.3\">entropy = 1.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3862.5\" y=\"-387.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3851.5\" y=\"-372.3\">value = [0, 4, 4]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3859.5\" y=\"-357.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 138&#45;&gt;140 -->\n",
       "<g class=\"edge\" id=\"edge140\">\n",
       "<title>138-&gt;140</title>\n",
       "<path d=\"M3815.8692,-460.8796C3828.0148,-449.2237 3841.2119,-436.5587 3853.3451,-424.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3855.9516,-427.2643 3860.7433,-417.8149 3851.1047,-422.2138 3855.9516,-427.2643\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 143 -->\n",
       "<g class=\"node\" id=\"node144\">\n",
       "<title>143</title>\n",
       "<polygon fill=\"#61ea9a\" points=\"4024,-782 3911,-782 3911,-699 4024,-699 4024,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3933.5\" y=\"-766.8\">x_1 ≤ 0.711</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3926\" y=\"-751.8\">entropy = 0.65</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3930\" y=\"-736.8\">samples = 24</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3919\" y=\"-721.8\">value = [4, 20, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3930.5\" y=\"-706.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 142&#45;&gt;143 -->\n",
       "<g class=\"edge\" id=\"edge143\">\n",
       "<title>142-&gt;143</title>\n",
       "<path d=\"M3967.5,-817.8796C3967.5,-809.6838 3967.5,-800.9891 3967.5,-792.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3971.0001,-792.298 3967.5,-782.2981 3964.0001,-792.2981 3971.0001,-792.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 146 -->\n",
       "<g class=\"node\" id=\"node147\">\n",
       "<title>146</title>\n",
       "<polygon fill=\"#eda775\" points=\"4223.5,-782 4103.5,-782 4103.5,-699 4223.5,-699 4223.5,-782\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4129.5\" y=\"-766.8\">x_1 ≤ 0.482</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4118.5\" y=\"-751.8\">entropy = 0.785</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4126\" y=\"-736.8\">samples = 77</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4111.5\" y=\"-721.8\">value = [59, 18, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4133.5\" y=\"-706.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 142&#45;&gt;146 -->\n",
       "<g class=\"edge\" id=\"edge146\">\n",
       "<title>142-&gt;146</title>\n",
       "<path d=\"M4027.6385,-822.9874C4048.8977,-810.0799 4072.966,-795.4671 4094.8758,-782.1647\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"4096.7438,-785.1252 4103.4753,-776.9436 4093.1109,-779.1417 4096.7438,-785.1252\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 144 -->\n",
       "<g class=\"node\" id=\"node145\">\n",
       "<title>144</title>\n",
       "<polygon fill=\"#4be78c\" points=\"3964,-655.5 3851,-655.5 3851,-587.5 3964,-587.5 3964,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3862.5\" y=\"-640.3\">entropy = 0.414</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3870\" y=\"-625.3\">samples = 12</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3859\" y=\"-610.3\">value = [1, 11, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3870.5\" y=\"-595.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 143&#45;&gt;144 -->\n",
       "<g class=\"edge\" id=\"edge144\">\n",
       "<title>143-&gt;144</title>\n",
       "<path d=\"M3946.5149,-698.8796C3940.9707,-687.8835 3934.9736,-675.9893 3929.3831,-664.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"3932.4291,-663.1684 3924.8016,-655.8149 3926.1786,-666.3199 3932.4291,-663.1684\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 145 -->\n",
       "<g class=\"node\" id=\"node146\">\n",
       "<title>145</title>\n",
       "<polygon fill=\"#7beeab\" points=\"4088.5,-655.5 3982.5,-655.5 3982.5,-587.5 4088.5,-587.5 4088.5,-655.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3990.5\" y=\"-640.3\">entropy = 0.811</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3998\" y=\"-625.3\">samples = 12</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3990.5\" y=\"-610.3\">value = [3, 9, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3998.5\" y=\"-595.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 143&#45;&gt;145 -->\n",
       "<g class=\"edge\" id=\"edge145\">\n",
       "<title>143-&gt;145</title>\n",
       "<path d=\"M3991.2831,-698.8796C3997.6294,-687.7735 4004.4993,-675.7513 4010.8891,-664.5691\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"4013.9689,-666.2339 4015.8915,-655.8149 4007.8912,-662.7609 4013.9689,-666.2339\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 147 -->\n",
       "<g class=\"node\" id=\"node148\">\n",
       "<title>147</title>\n",
       "<polygon fill=\"#93f1ba\" points=\"4220,-663 4107,-663 4107,-580 4220,-580 4220,-663\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4129.5\" y=\"-647.8\">x_1 ≤ 0.425</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4118.5\" y=\"-632.8\">entropy = 0.896</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4126\" y=\"-617.8\">samples = 16</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4115\" y=\"-602.8\">value = [5, 11, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4126.5\" y=\"-587.8\">class = white</text>\n",
       "</g>\n",
       "<!-- 146&#45;&gt;147 -->\n",
       "<g class=\"edge\" id=\"edge147\">\n",
       "<title>146-&gt;147</title>\n",
       "<path d=\"M4163.5,-698.8796C4163.5,-690.6838 4163.5,-681.9891 4163.5,-673.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"4167.0001,-673.298 4163.5,-663.2981 4160.0001,-673.2981 4167.0001,-673.298\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 150 -->\n",
       "<g class=\"node\" id=\"node151\">\n",
       "<title>150</title>\n",
       "<polygon fill=\"#e89153\" points=\"4351,-663 4238,-663 4238,-580 4351,-580 4351,-663\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4260.5\" y=\"-647.8\">x_1 ≤ 0.753</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4249.5\" y=\"-632.8\">entropy = 0.514</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4257\" y=\"-617.8\">samples = 61</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4246\" y=\"-602.8\">value = [54, 7, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4264.5\" y=\"-587.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 146&#45;&gt;150 -->\n",
       "<g class=\"edge\" id=\"edge150\">\n",
       "<title>146-&gt;150</title>\n",
       "<path d=\"M4209.3174,-698.8796C4219.5294,-689.6031 4230.445,-679.6874 4240.9376,-670.1559\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"4243.4384,-672.6127 4248.487,-663.2981 4238.7316,-667.4313 4243.4384,-672.6127\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 148 -->\n",
       "<g class=\"node\" id=\"node149\">\n",
       "<title>148</title>\n",
       "<polygon fill=\"#7beeab\" points=\"4093.5,-536.5 3987.5,-536.5 3987.5,-468.5 4093.5,-468.5 4093.5,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3995.5\" y=\"-521.3\">entropy = 0.811</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4006.5\" y=\"-506.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"3995.5\" y=\"-491.3\">value = [2, 6, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4003.5\" y=\"-476.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 147&#45;&gt;148 -->\n",
       "<g class=\"edge\" id=\"edge148\">\n",
       "<title>147-&gt;148</title>\n",
       "<path d=\"M4120.4806,-579.8796C4108.4329,-568.2237 4095.3422,-555.5587 4083.3069,-543.9148\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"4085.589,-541.2528 4075.9684,-536.8149 4080.7217,-546.2837 4085.589,-541.2528\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 149 -->\n",
       "<g class=\"node\" id=\"node150\">\n",
       "<title>149</title>\n",
       "<polygon fill=\"#b0f5cd\" points=\"4217.5,-536.5 4111.5,-536.5 4111.5,-468.5 4217.5,-468.5 4217.5,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4119.5\" y=\"-521.3\">entropy = 0.954</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4130.5\" y=\"-506.3\">samples = 8</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4119.5\" y=\"-491.3\">value = [3, 5, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4127.5\" y=\"-476.3\">class = white</text>\n",
       "</g>\n",
       "<!-- 147&#45;&gt;149 -->\n",
       "<g class=\"edge\" id=\"edge149\">\n",
       "<title>147-&gt;149</title>\n",
       "<path d=\"M4163.8498,-579.8796C4163.9394,-569.2134 4164.0361,-557.7021 4164.1269,-546.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"4167.6274,-546.844 4164.2116,-536.8149 4160.6276,-546.7851 4167.6274,-546.844\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 151 -->\n",
       "<g class=\"node\" id=\"node152\">\n",
       "<title>151</title>\n",
       "<polygon fill=\"#eb9f69\" points=\"4350,-544 4237,-544 4237,-461 4350,-461 4350,-544\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4259.5\" y=\"-528.8\">x_2 ≤ 0.598</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4248.5\" y=\"-513.8\">entropy = 0.711</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4256\" y=\"-498.8\">samples = 36</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4245\" y=\"-483.8\">value = [29, 7, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4263.5\" y=\"-468.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 150&#45;&gt;151 -->\n",
       "<g class=\"edge\" id=\"edge151\">\n",
       "<title>150-&gt;151</title>\n",
       "<path d=\"M4294.1502,-579.8796C4294.0814,-571.6838 4294.0083,-562.9891 4293.937,-554.5013\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"4297.4352,-554.2683 4293.8512,-544.2981 4290.4355,-554.3272 4297.4352,-554.2683\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 156 -->\n",
       "<g class=\"node\" id=\"node157\">\n",
       "<title>156</title>\n",
       "<polygon fill=\"#e58139\" points=\"4481,-536.5 4368,-536.5 4368,-468.5 4481,-468.5 4481,-536.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4386.5\" y=\"-521.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4387\" y=\"-506.3\">samples = 25</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4376\" y=\"-491.3\">value = [25, 0, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4394.5\" y=\"-476.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 150&#45;&gt;156 -->\n",
       "<g class=\"edge\" id=\"edge156\">\n",
       "<title>150-&gt;156</title>\n",
       "<path d=\"M4339.9677,-579.8796C4352.8211,-568.1138 4366.7979,-555.3197 4379.6166,-543.5855\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"4382.0001,-546.1488 4387.0131,-536.8149 4377.2736,-540.9854 4382.0001,-546.1488\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 152 -->\n",
       "<g class=\"node\" id=\"node153\">\n",
       "<title>152</title>\n",
       "<polygon fill=\"#fae6d7\" points=\"4282.5,-417.5 4176.5,-417.5 4176.5,-349.5 4282.5,-349.5 4282.5,-417.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4184.5\" y=\"-402.3\">entropy = 0.991</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4195.5\" y=\"-387.3\">samples = 9</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4184.5\" y=\"-372.3\">value = [5, 4, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4199.5\" y=\"-357.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 151&#45;&gt;152 -->\n",
       "<g class=\"edge\" id=\"edge152\">\n",
       "<title>151-&gt;152</title>\n",
       "<path d=\"M4271.1159,-460.8796C4265.202,-449.8835 4258.8051,-437.9893 4252.842,-426.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"4255.7742,-424.9642 4247.9551,-417.8149 4249.6092,-428.2799 4255.7742,-424.9642\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 153 -->\n",
       "<g class=\"node\" id=\"node154\">\n",
       "<title>153</title>\n",
       "<polygon fill=\"#e89152\" points=\"4414,-425 4301,-425 4301,-342 4414,-342 4414,-425\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4323.5\" y=\"-409.8\">x_1 ≤ 0.544</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4312.5\" y=\"-394.8\">entropy = 0.503</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4320\" y=\"-379.8\">samples = 27</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4309\" y=\"-364.8\">value = [24, 3, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4327.5\" y=\"-349.8\">class = red</text>\n",
       "</g>\n",
       "<!-- 151&#45;&gt;153 -->\n",
       "<g class=\"edge\" id=\"edge153\">\n",
       "<title>151-&gt;153</title>\n",
       "<path d=\"M4315.8841,-460.8796C4320.4856,-452.3236 4325.3796,-443.2238 4330.1342,-434.3833\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"4333.3662,-435.763 4335.0204,-425.2981 4327.2012,-432.4473 4333.3662,-435.763\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 154 -->\n",
       "<g class=\"node\" id=\"node155\">\n",
       "<title>154</title>\n",
       "<polygon fill=\"#f0b78e\" points=\"4346.5,-298.5 4240.5,-298.5 4240.5,-230.5 4346.5,-230.5 4346.5,-298.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4248.5\" y=\"-283.3\">entropy = 0.881</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4256\" y=\"-268.3\">samples = 10</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4248.5\" y=\"-253.3\">value = [7, 3, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4263.5\" y=\"-238.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 153&#45;&gt;154 -->\n",
       "<g class=\"edge\" id=\"edge154\">\n",
       "<title>153-&gt;154</title>\n",
       "<path d=\"M4335.1159,-341.8796C4329.202,-330.8835 4322.8051,-318.9893 4316.842,-307.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"4319.7742,-305.9642 4311.9551,-298.8149 4313.6092,-309.2799 4319.7742,-305.9642\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "<!-- 155 -->\n",
       "<g class=\"node\" id=\"node156\">\n",
       "<title>155</title>\n",
       "<polygon fill=\"#e58139\" points=\"4478,-298.5 4365,-298.5 4365,-230.5 4478,-230.5 4478,-298.5\" stroke=\"#000000\"/>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4383.5\" y=\"-283.3\">entropy = 0.0</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4384\" y=\"-268.3\">samples = 17</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4373\" y=\"-253.3\">value = [17, 0, 0]</text>\n",
       "<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"start\" x=\"4391.5\" y=\"-238.3\">class = red</text>\n",
       "</g>\n",
       "<!-- 153&#45;&gt;155 -->\n",
       "<g class=\"edge\" id=\"edge155\">\n",
       "<title>153-&gt;155</title>\n",
       "<path d=\"M4379.8841,-341.8796C4385.798,-330.8835 4392.1949,-318.9893 4398.158,-307.9015\" fill=\"none\" stroke=\"#000000\"/>\n",
       "<polygon fill=\"#000000\" points=\"4401.3908,-309.2799 4403.0449,-298.8149 4395.2258,-305.9642 4401.3908,-309.2799\" stroke=\"#000000\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dot_data = tree.export_graphviz(model2Norway, out_file=None, \n",
    "                    feature_names=['x_1','x_2'],  \n",
    "                    class_names=norwayColors,  \n",
    "                    filled=True, rounded=False,  \n",
    "                    special_characters=True)  \n",
    "graph = graphviz.Source(dot_data)  \n",
    "display(SVG(graph.pipe(format='svg')))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conclusion\n",
    "\n",
    "Tree classifier provides a simple algorithm that is able to cope with non linearly separable problems like Norway flag.\n",
    "\n",
    "Its performance, as measured by the F1-score or accuracy, is very close to the one achieved by neural networks (see [HTML](ClassificationMulti2Features-Keras.html) / [Jupyter](ClassificationMulti2Features-Keras.ipynb)) but the design cost is way lower. The prediction cost might also be lower as well : it requires maximum $d = log_2(N)$ comparisons where $d$ is the depth and $N$ the number of leaves of the tree."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Where to go from here\n",
    "\n",
    "- Same multi-class problem but using neural networks and the Keras framework ([HTML](ClassificationMulti2Features-Keras.html) / [Jupyter](ClassificationMulti2Features-Keras.ipynb))\n",
    "- Binary classification with Logistic regression ([HTML](ClassificationContinuous2Features-KNN.html) / [Jupyter](ClassificationContinuous2Features.ipynb)), or k Nearest Neighbors ([HTML](ClassificationContinuous2Features.html) / [Jupyter](ClassificationContinuous2Features-KNN.ipynb))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}