misc/example/03_model_example_bace_dual_path.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from molmap.model import RegressionEstimator, MultiClassEstimator, MultiLabelEstimator\n",
    "from sklearn.preprocessing import StandardScaler, MinMaxScaler\n",
    "from chembench import dataset\n",
    "from sklearn.utils import shuffle \n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from molmap import MolMap\n",
    "from molmap import feature\n",
    "\n",
    "def Rdsplit(df, random_state = 888, split_size = [0.8, 0.1, 0.1]):\n",
    "    base_indices = np.arange(len(df)) \n",
    "    base_indices = shuffle(base_indices, random_state = random_state) \n",
    "    nb_test = int(len(base_indices) * split_size[2]) \n",
    "    nb_val = int(len(base_indices) * split_size[1]) \n",
    "    test_idx = base_indices[0:nb_test] \n",
    "    valid_idx = base_indices[(nb_test):(nb_test+nb_val)] \n",
    "    train_idx = base_indices[(nb_test+nb_val):len(base_indices)] \n",
    "    print(len(train_idx), len(valid_idx), len(test_idx)) \n",
    "    return train_idx, valid_idx, test_idx "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total samples: 1513\n"
     ]
    }
   ],
   "source": [
    "data = dataset.load_BACE()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pre-fit your molmap object "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-24 09:16:45,881 - \u001b[32mINFO\u001b[0m - [bidd-molmap]\u001b[0m - Applying grid feature map(assignment), this may take several minutes(1~30 min)\u001b[0m\n",
      "2021-04-24 09:16:48,708 - \u001b[32mINFO\u001b[0m - [bidd-molmap]\u001b[0m - Finished\u001b[0m\n"
     ]
    }
   ],
   "source": [
    "mp1 = MolMap(ftype='descriptor',metric='cosine',)\n",
    "mp1.fit(verbose=0, method='umap', min_dist=0.1, n_neighbors=15,)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-04-24 09:16:53,896 - \u001b[32mINFO\u001b[0m - [bidd-molmap]\u001b[0m - Applying naive scatter feature map...\u001b[0m\n",
      "2021-04-24 09:16:53,904 - \u001b[32mINFO\u001b[0m - [bidd-molmap]\u001b[0m - Finished\u001b[0m\n"
     ]
    }
   ],
   "source": [
    "bitsinfo = feature.fingerprint.Extraction().bitsinfo\n",
    "flist = bitsinfo[bitsinfo.Subtypes.isin(['PubChemFP'])].IDs.tolist()\n",
    "mp2 = MolMap(ftype = 'fingerprint', fmap_type = 'scatter', flist = flist) # \n",
    "mp2.fit(method = 'umap',  min_dist = 0.1, n_neighbors = 15, verbose = 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Extract Fmaps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|##########| 1513/1513 [01:30<00:00, 16.68it/s]\n",
      "100%|##########| 1513/1513 [00:49<00:00, 30.60it/s]\n"
     ]
    }
   ],
   "source": [
    "X1 = mp1.batch_transform(data.x)\n",
    "X2 = mp2.batch_transform(data.x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1513, 2)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y = pd.get_dummies(data.df['Class']).values\n",
    "Y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1211 151 151\n"
     ]
    }
   ],
   "source": [
    "train_idx, valid_idx, test_idx  = Rdsplit(data.df, random_state = 888)\n",
    "trainX = (X1[train_idx], X2[train_idx])\n",
    "validX = (X1[valid_idx], X2[valid_idx])\n",
    "testX = (X1[test_idx], X2[test_idx])\n",
    "\n",
    "trainY = Y[train_idx]\n",
    "validY = Y[valid_idx]\n",
    "testY = Y[test_idx]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MultiClassEstimator(gpuid='0')\n"
     ]
    }
   ],
   "source": [
    "# define your model, note that if your task is a multi-label problem, you should use MultiLabelEstimator\n",
    "clf = MultiClassEstimator(n_outputs=trainY.shape[1], \n",
    "                          fmap_shape1 = X1.shape[1:],\n",
    "                          fmap_shape2 = X2.shape[1:],\n",
    "                          metric='ROC', \n",
    "                          dense_layers = [128, 64],  gpuid = 0) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch: 0001, loss: 0.6898 - val_loss: 0.6794; auc: 0.7948 - val_auc: 0.7536                                                                                                    \n",
      "epoch: 0002, loss: 0.6759 - val_loss: 0.6679; auc: 0.8105 - val_auc: 0.7770                                                                                                    \n",
      "epoch: 0003, loss: 0.6627 - val_loss: 0.6561; auc: 0.8193 - val_auc: 0.7896                                                                                                    \n",
      "epoch: 0004, loss: 0.6497 - val_loss: 0.6459; auc: 0.8232 - val_auc: 0.7969                                                                                                    \n",
      "epoch: 0005, loss: 0.6418 - val_loss: 0.6328; auc: 0.8292 - val_auc: 0.8012                                                                                                    \n",
      "epoch: 0006, loss: 0.6276 - val_loss: 0.6213; auc: 0.8391 - val_auc: 0.8125                                                                                                    \n",
      "epoch: 0007, loss: 0.6071 - val_loss: 0.6040; auc: 0.8414 - val_auc: 0.8154                                                                                                    \n",
      "epoch: 0008, loss: 0.5869 - val_loss: 0.5844; auc: 0.8478 - val_auc: 0.8205                                                                                                    \n",
      "epoch: 0009, loss: 0.5658 - val_loss: 0.5686; auc: 0.8523 - val_auc: 0.8266                                                                                                    \n",
      "epoch: 0010, loss: 0.5520 - val_loss: 0.5640; auc: 0.8548 - val_auc: 0.8260                                                                                                    \n",
      "epoch: 0011, loss: 0.5339 - val_loss: 0.5502; auc: 0.8596 - val_auc: 0.8317                                                                                                    \n",
      "epoch: 0012, loss: 0.5144 - val_loss: 0.5314; auc: 0.8646 - val_auc: 0.8335                                                                                                    \n",
      "epoch: 0013, loss: 0.5047 - val_loss: 0.5201; auc: 0.8686 - val_auc: 0.8378                                                                                                    \n",
      "epoch: 0014, loss: 0.4945 - val_loss: 0.5211; auc: 0.8724 - val_auc: 0.8381                                                                                                    \n",
      "epoch: 0015, loss: 0.4858 - val_loss: 0.5123; auc: 0.8753 - val_auc: 0.8410                                                                                                    \n",
      "epoch: 0016, loss: 0.4709 - val_loss: 0.5024; auc: 0.8795 - val_auc: 0.8422                                                                                                    \n",
      "epoch: 0017, loss: 0.4692 - val_loss: 0.5098; auc: 0.8812 - val_auc: 0.8429                                                                                                    \n",
      "epoch: 0018, loss: 0.4623 - val_loss: 0.5035; auc: 0.8848 - val_auc: 0.8463                                                                                                    \n",
      "epoch: 0019, loss: 0.4575 - val_loss: 0.5092; auc: 0.8875 - val_auc: 0.8468                                                                                                    \n",
      "epoch: 0020, loss: 0.4575 - val_loss: 0.5199; auc: 0.8903 - val_auc: 0.8484                                                                                                    \n",
      "epoch: 0021, loss: 0.4655 - val_loss: 0.4843; auc: 0.8923 - val_auc: 0.8513                                                                                                    \n",
      "epoch: 0022, loss: 0.4339 - val_loss: 0.4853; auc: 0.8945 - val_auc: 0.8507                                                                                                    \n",
      "epoch: 0023, loss: 0.4277 - val_loss: 0.4833; auc: 0.8965 - val_auc: 0.8500                                                                                                    \n",
      "epoch: 0024, loss: 0.4199 - val_loss: 0.4879; auc: 0.8986 - val_auc: 0.8513                                                                                                    \n",
      "epoch: 0025, loss: 0.4228 - val_loss: 0.4774; auc: 0.9019 - val_auc: 0.8511                                                                                                    \n",
      "epoch: 0026, loss: 0.4090 - val_loss: 0.4763; auc: 0.9038 - val_auc: 0.8515                                                                                                    \n",
      "epoch: 0027, loss: 0.4126 - val_loss: 0.4766; auc: 0.9055 - val_auc: 0.8522                                                                                                    \n",
      "epoch: 0028, loss: 0.4083 - val_loss: 0.4742; auc: 0.9073 - val_auc: 0.8518                                                                                                    \n",
      "epoch: 0029, loss: 0.4071 - val_loss: 0.4967; auc: 0.9094 - val_auc: 0.8539                                                                                                    \n",
      "epoch: 0030, loss: 0.4039 - val_loss: 0.4930; auc: 0.9111 - val_auc: 0.8543                                                                                                    \n",
      "epoch: 0031, loss: 0.4027 - val_loss: 0.4749; auc: 0.9122 - val_auc: 0.8548                                                                                                    \n",
      "epoch: 0032, loss: 0.3928 - val_loss: 0.4736; auc: 0.9142 - val_auc: 0.8548                                                                                                    \n",
      "epoch: 0033, loss: 0.3846 - val_loss: 0.4748; auc: 0.9164 - val_auc: 0.8573                                                                                                    \n",
      "epoch: 0034, loss: 0.3906 - val_loss: 0.4801; auc: 0.9173 - val_auc: 0.8575                                                                                                    \n",
      "epoch: 0035, loss: 0.3836 - val_loss: 0.4688; auc: 0.9192 - val_auc: 0.8564                                                                                                    \n",
      "epoch: 0036, loss: 0.3734 - val_loss: 0.4740; auc: 0.9210 - val_auc: 0.8584                                                                                                    \n",
      "epoch: 0037, loss: 0.3757 - val_loss: 0.4779; auc: 0.9218 - val_auc: 0.8591                                                                                                    \n",
      "epoch: 0038, loss: 0.3738 - val_loss: 0.4891; auc: 0.9223 - val_auc: 0.8579                                                                                                    \n",
      "epoch: 0039, loss: 0.3828 - val_loss: 0.4664; auc: 0.9246 - val_auc: 0.8595                                                                                                    \n",
      "epoch: 0040, loss: 0.3668 - val_loss: 0.4642; auc: 0.9258 - val_auc: 0.8579                                                                                                    \n",
      "epoch: 0041, loss: 0.3580 - val_loss: 0.4766; auc: 0.9273 - val_auc: 0.8603                                                                                                    \n",
      "epoch: 0042, loss: 0.3563 - val_loss: 0.4670; auc: 0.9289 - val_auc: 0.8605                                                                                                    \n",
      "epoch: 0043, loss: 0.3523 - val_loss: 0.4618; auc: 0.9303 - val_auc: 0.8618                                                                                                    \n",
      "epoch: 0044, loss: 0.3501 - val_loss: 0.4609; auc: 0.9317 - val_auc: 0.8632                                                                                                    \n",
      "epoch: 0045, loss: 0.3433 - val_loss: 0.4731; auc: 0.9326 - val_auc: 0.8651                                                                                                    \n",
      "epoch: 0046, loss: 0.3505 - val_loss: 0.4637; auc: 0.9340 - val_auc: 0.8634                                                                                                    \n",
      "epoch: 0047, loss: 0.3407 - val_loss: 0.4622; auc: 0.9355 - val_auc: 0.8643                                                                                                    \n",
      "epoch: 0048, loss: 0.3407 - val_loss: 0.4601; auc: 0.9370 - val_auc: 0.8658                                                                                                    \n",
      "epoch: 0049, loss: 0.3341 - val_loss: 0.4782; auc: 0.9378 - val_auc: 0.8658                                                                                                    \n",
      "epoch: 0050, loss: 0.3417 - val_loss: 0.4606; auc: 0.9389 - val_auc: 0.8673                                                                                                    \n",
      "epoch: 0051, loss: 0.3276 - val_loss: 0.4704; auc: 0.9406 - val_auc: 0.8682                                                                                                    \n",
      "epoch: 0052, loss: 0.3323 - val_loss: 0.4664; auc: 0.9415 - val_auc: 0.8683                                                                                                    \n",
      "epoch: 0053, loss: 0.3210 - val_loss: 0.4892; auc: 0.9429 - val_auc: 0.8676                                                                                                    \n",
      "epoch: 0054, loss: 0.3435 - val_loss: 0.4928; auc: 0.9437 - val_auc: 0.8682                                                                                                    \n",
      "epoch: 0055, loss: 0.3376 - val_loss: 0.4731; auc: 0.9437 - val_auc: 0.8701                                                                                                    \n",
      "epoch: 0056, loss: 0.3151 - val_loss: 0.4705; auc: 0.9447 - val_auc: 0.8701                                                                                                    \n",
      "epoch: 0057, loss: 0.3388 - val_loss: 0.4658; auc: 0.9463 - val_auc: 0.8676                                                                                                    \n",
      "epoch: 0058, loss: 0.3201 - val_loss: 0.4737; auc: 0.9475 - val_auc: 0.8712                                                                                                    \n",
      "epoch: 0059, loss: 0.3241 - val_loss: 0.4959; auc: 0.9468 - val_auc: 0.8671                                                                                                    \n",
      "epoch: 0060, loss: 0.3189 - val_loss: 0.4713; auc: 0.9482 - val_auc: 0.8710                                                                                                    \n",
      "epoch: 0061, loss: 0.3271 - val_loss: 0.4902; auc: 0.9499 - val_auc: 0.8710                                                                                                    \n",
      "epoch: 0062, loss: 0.3172 - val_loss: 0.4627; auc: 0.9510 - val_auc: 0.8728                                                                                                    \n",
      "epoch: 0063, loss: 0.3127 - val_loss: 0.4847; auc: 0.9519 - val_auc: 0.8706                                                                                                    \n",
      "epoch: 0064, loss: 0.3016 - val_loss: 0.4688; auc: 0.9534 - val_auc: 0.8719                                                                                                    \n",
      "epoch: 0065, loss: 0.3167 - val_loss: 0.4666; auc: 0.9534 - val_auc: 0.8694                                                                                                    \n",
      "epoch: 0066, loss: 0.3013 - val_loss: 0.4854; auc: 0.9554 - val_auc: 0.8728                                                                                                    \n",
      "epoch: 0067, loss: 0.3041 - val_loss: 0.4623; auc: 0.9557 - val_auc: 0.8703                                                                                                    \n",
      "epoch: 0068, loss: 0.2875 - val_loss: 0.4611; auc: 0.9558 - val_auc: 0.8737                                                                                                    \n",
      "epoch: 0069, loss: 0.2905 - val_loss: 0.4632; auc: 0.9570 - val_auc: 0.8733                                                                                                    \n",
      "epoch: 0070, loss: 0.2860 - val_loss: 0.4739; auc: 0.9583 - val_auc: 0.8735                                                                                                    \n",
      "epoch: 0071, loss: 0.2859 - val_loss: 0.4662; auc: 0.9580 - val_auc: 0.8717                                                                                                    \n",
      "epoch: 0072, loss: 0.2767 - val_loss: 0.4613; auc: 0.9598 - val_auc: 0.8742                                                                                                    \n",
      "epoch: 0073, loss: 0.2772 - val_loss: 0.4681; auc: 0.9603 - val_auc: 0.8722                                                                                                    \n",
      "epoch: 0074, loss: 0.2705 - val_loss: 0.4667; auc: 0.9613 - val_auc: 0.8740                                                                                                    \n",
      "epoch: 0075, loss: 0.2707 - val_loss: 0.4665; auc: 0.9625 - val_auc: 0.8730                                                                                                    \n",
      "epoch: 0076, loss: 0.2694 - val_loss: 0.4903; auc: 0.9626 - val_auc: 0.8728                                                                                                    \n",
      "epoch: 0077, loss: 0.2702 - val_loss: 0.4811; auc: 0.9635 - val_auc: 0.8738                                                                                                    \n",
      "epoch: 0078, loss: 0.2701 - val_loss: 0.4692; auc: 0.9642 - val_auc: 0.8742                                                                                                    \n",
      "epoch: 0079, loss: 0.2667 - val_loss: 0.4818; auc: 0.9638 - val_auc: 0.8715                                                                                                    \n",
      "epoch: 0080, loss: 0.2604 - val_loss: 0.4701; auc: 0.9659 - val_auc: 0.8753                                                                                                    \n",
      "epoch: 0081, loss: 0.2569 - val_loss: 0.4905; auc: 0.9664 - val_auc: 0.8749                                                                                                    \n",
      "epoch: 0082, loss: 0.2569 - val_loss: 0.4717; auc: 0.9672 - val_auc: 0.8751                                                                                                    \n",
      "epoch: 0083, loss: 0.2701 - val_loss: 0.4768; auc: 0.9680 - val_auc: 0.8754                                                                                                    \n",
      "epoch: 0084, loss: 0.2519 - val_loss: 0.4756; auc: 0.9689 - val_auc: 0.8740                                                                                                    \n",
      "epoch: 0085, loss: 0.2485 - val_loss: 0.4793; auc: 0.9694 - val_auc: 0.8730                                                                                                    \n",
      "epoch: 0086, loss: 0.2590 - val_loss: 0.4872; auc: 0.9694 - val_auc: 0.8735                                                                                                    \n",
      "epoch: 0087, loss: 0.2526 - val_loss: 0.4749; auc: 0.9705 - val_auc: 0.8772                                                                                                    \n",
      "epoch: 0088, loss: 0.2567 - val_loss: 0.5360; auc: 0.9704 - val_auc: 0.8753                                                                                                    \n",
      "epoch: 0089, loss: 0.2800 - val_loss: 0.5434; auc: 0.9703 - val_auc: 0.8738                                                                                                    \n",
      "epoch: 0090, loss: 0.2855 - val_loss: 0.5142; auc: 0.9700 - val_auc: 0.8749                                                                                                    \n",
      "epoch: 0091, loss: 0.2591 - val_loss: 0.4916; auc: 0.9710 - val_auc: 0.8744                                                                                                    \n",
      "epoch: 0092, loss: 0.2561 - val_loss: 0.4809; auc: 0.9725 - val_auc: 0.8746                                                                                                    \n",
      "epoch: 0093, loss: 0.2392 - val_loss: 0.4918; auc: 0.9729 - val_auc: 0.8765                                                                                                    \n",
      "epoch: 0094, loss: 0.2411 - val_loss: 0.4984; auc: 0.9733 - val_auc: 0.8758                                                                                                    \n",
      "epoch: 0095, loss: 0.2442 - val_loss: 0.4775; auc: 0.9750 - val_auc: 0.8774                                                                                                    \n",
      "epoch: 0096, loss: 0.2415 - val_loss: 0.4857; auc: 0.9732 - val_auc: 0.8735                                                                                                    \n",
      "epoch: 0097, loss: 0.2418 - val_loss: 0.4931; auc: 0.9740 - val_auc: 0.8742                                                                                                    \n",
      "epoch: 0098, loss: 0.2260 - val_loss: 0.5066; auc: 0.9753 - val_auc: 0.8747                                                                                                    \n",
      "\n",
      "Restoring model weights from the end of the best epoch.\n",
      "\n",
      "Epoch 00098: early stopping\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "MultiClassEstimator(gpuid='0')"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fit your model\n",
    "clf.fit(trainX, trainY, validX, validY )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## plot training history"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pd.DataFrame(clf.history.history)[['accuracy', 'val_accuracy']].rename(columns={'accuracy':'auc', 'val_accuracy':'val_auc'}).plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pd.DataFrame(clf.history.history)[['loss', 'val_loss']].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best epochs: 47.00, Best loss: 0.46\n"
     ]
    }
   ],
   "source": [
    "print('Best epochs: %.2f, Best loss: %.2f' % (clf._performance.best_epoch, clf._performance.best))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# performance on  test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "auc = clf.score(testX, testY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8942959001782531"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "auc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_pred = pd.DataFrame([testY[:, 0], clf.predict_proba(testX)[:,0]]).T\n",
    "df_pred.columns=['y_true', 'y_pred_prob']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>y_true</th>\n",
       "      <th>y_pred_prob</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.959249</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.421290</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.967951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.875520</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.137767</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   y_true  y_pred_prob\n",
       "0     1.0     0.959249\n",
       "1     0.0     0.421290\n",
       "2     1.0     0.967951\n",
       "3     1.0     0.875520\n",
       "4     0.0     0.137767"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_pred.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_pred.plot(figsize=(16, 4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>y_true</th>\n",
       "      <th>y_pred_prob</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>y_true</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.689067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>y_pred_prob</th>\n",
       "      <td>0.689067</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               y_true  y_pred_prob\n",
       "y_true       1.000000     0.689067\n",
       "y_pred_prob  0.689067     1.000000"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_pred.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}