Dataset_ComplexityEstimation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Repositories uploaded!!\n"
     ]
    }
   ],
   "source": [
    "# from two_TrAdaBoostR2 import TwoStageTrAdaBoostR2 ##STrAdaBoost.R2\n",
    "# from TwoStageTrAdaBoostR2 import TwoStageTrAdaBoostR2 ##two-stage TrAdaBoost.R2\n",
    "\n",
    "import pandas as pd\n",
    "import sys\n",
    "import numpy as np\n",
    "from pandas import DataFrame\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.ensemble import AdaBoostRegressor\n",
    "\n",
    "from keras.models import Sequential, load_model, Model\n",
    "from keras.layers import Input, Dense, Activation, Conv2D, Dropout, Flatten\n",
    "from keras import optimizers, utils, initializers, regularizers\n",
    "import keras.backend as K\n",
    "\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.preprocessing import StandardScaler #Importing the StandardScaler\n",
    "\n",
    "from itertools import combinations\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from scipy.stats.stats import pearsonr\n",
    "from math import sqrt\n",
    "\n",
    "#Geo plotting libraries\n",
    "import geopandas as gdp\n",
    "from matplotlib.colors import ListedColormap\n",
    "# import geoplot as glpt\n",
    "\n",
    "import xgboost as xgb\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.linear_model import Ridge\n",
    "from sklearn import linear_model\n",
    "from keras.models import Sequential\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "from sklearn.ensemble import GradientBoostingRegressor\n",
    "\n",
    "from sklearn.model_selection import KFold\n",
    "import matplotlib.lines as mlines\n",
    "\n",
    "import statistics \n",
    "\n",
    "from scipy.stats import rv_continuous\n",
    "from scipy.stats import *\n",
    "\n",
    "from statistics import mean\n",
    "from sklearn.cluster import KMeans\n",
    "from scipy.spatial import distance\n",
    "\n",
    "\n",
    "######### Instance Transfer repositories ####################\n",
    "from adapt.instance_based import TwoStageTrAdaBoostR2\n",
    "\n",
    "print(\"Repositories uploaded!!\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Second Upload Completed!!\n"
     ]
    }
   ],
   "source": [
    "from adapt.instance_based import TrAdaBoost, TrAdaBoostR2, TwoStageTrAdaBoostR2\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from adapt.instance_based import KMM\n",
    "\n",
    "print(\"Second Upload Completed!!\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Concrete Data\n",
      "-------------------------------------------\n",
      "(1030, 9)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Index(['Cement', 'BlastFurnaceSlag', 'FlyAsh', 'Water', 'Superplasticizer',\n",
       "       'CoarseAggregate', 'FineAggregate', 'Age',\n",
       "       'ConcreteCompressiveStrength'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 222,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "################################### Concrete ###########################################################################################################\n",
    "target_column_concrete = ['ConcreteCompressiveStrength']\n",
    "ConcreteData_df = pd.read_excel('UCI_regression/Concrete/Concrete_Data.xls') ## 'Cement' found to be correlated at 0.4 :: 100\n",
    "print(\"Concrete Data\")\n",
    "print(\"-------------------------------------------\")\n",
    "print(ConcreteData_df.shape)\n",
    "\n",
    "concrete_df_y = ConcreteData_df[target_column_concrete]\n",
    "concrete_df_X = ConcreteData_df.drop(target_column_concrete, axis = 1)\n",
    "\n",
    "ConcreteData_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([6.86922769, 6.16753072, 6.09706277, 6.93037381, 6.38141958,\n",
       "       6.93411207, 6.93187261, 6.33603673, 6.8252345 ])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get Entropy Concrete\n",
    "entropy_concrete = entropy(ConcreteData_df)\n",
    "entropy_concrete"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.3703907947964364"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Concrete\n",
    "corr_concrete =  ConcreteData_df.corr(method = 'spearman')['ConcreteCompressiveStrength'].abs()\n",
    "corr_concrete = corr_concrete.sort_values(ascending = True)\n",
    "corr_concrete\n",
    "\n",
    "sumval_concrete = corr_concrete.sum()/len(corr_concrete)\n",
    "sumval_concrete\n",
    "\n",
    "## 0% simple (>0.8)\n",
    "## 0% simple (>0.6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.10444626994175524"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Concrete for the features\n",
    "corr_concrete =  concrete_df_X.corr(method = 'spearman')\n",
    "corr_concrete\n",
    "\n",
    "corr_concrete_lsum = np.tril(corr_concrete).sum() - np.trace(corr_concrete)\n",
    "concrete_elems = len(corr_concrete) * (len(corr_concrete) - 1) / 2\n",
    "corr_concrete_lavg = abs(corr_concrete_lsum / concrete_elems)\n",
    "corr_concrete_lavg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8.214899392828647"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the L1 parameter Concrete\n",
    "from sklearn import linear_model\n",
    "\n",
    "regr_concrete = linear_model.LinearRegression()\n",
    "regr_concrete.fit(concrete_df_X, concrete_df_y)\n",
    "predict_concrete = regr_concrete.predict(concrete_df_X)\n",
    "\n",
    "concrete_df_y_list = concrete_df_y.to_numpy()\n",
    "mandis_concrete = np.abs(concrete_df_y_list - predict_concrete)\n",
    "L1_concrete = np.sum(mandis_concrete)/len(mandis_concrete)\n",
    "L1_concrete\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>BlastFurnaceSlag</th>\n",
       "      <th>FlyAsh</th>\n",
       "      <th>Water</th>\n",
       "      <th>Superplasticizer</th>\n",
       "      <th>CoarseAggregate</th>\n",
       "      <th>FineAggregate</th>\n",
       "      <th>Age</th>\n",
       "      <th>ConcreteCompressiveStrength</th>\n",
       "      <th>Residuals</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>162.0</td>\n",
       "      <td>2.5</td>\n",
       "      <td>1040.0</td>\n",
       "      <td>676.0</td>\n",
       "      <td>28</td>\n",
       "      <td>79.986111</td>\n",
       "      <td>1.631719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>162.0</td>\n",
       "      <td>2.5</td>\n",
       "      <td>1055.0</td>\n",
       "      <td>676.0</td>\n",
       "      <td>28</td>\n",
       "      <td>61.887366</td>\n",
       "      <td>0.378786</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>132.4</td>\n",
       "      <td>0.0</td>\n",
       "      <td>192.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>978.4</td>\n",
       "      <td>825.5</td>\n",
       "      <td>360</td>\n",
       "      <td>44.296075</td>\n",
       "      <td>1.038994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>114.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>228.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>932.0</td>\n",
       "      <td>670.0</td>\n",
       "      <td>90</td>\n",
       "      <td>47.029847</td>\n",
       "      <td>0.857170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>95.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>228.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>932.0</td>\n",
       "      <td>594.0</td>\n",
       "      <td>28</td>\n",
       "      <td>36.447770</td>\n",
       "      <td>-0.499601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1025</th>\n",
       "      <td>116.0</td>\n",
       "      <td>90.3</td>\n",
       "      <td>179.6</td>\n",
       "      <td>8.9</td>\n",
       "      <td>870.1</td>\n",
       "      <td>768.3</td>\n",
       "      <td>28</td>\n",
       "      <td>44.284354</td>\n",
       "      <td>0.610542</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1026</th>\n",
       "      <td>0.0</td>\n",
       "      <td>115.6</td>\n",
       "      <td>196.0</td>\n",
       "      <td>10.4</td>\n",
       "      <td>817.9</td>\n",
       "      <td>813.4</td>\n",
       "      <td>28</td>\n",
       "      <td>31.178794</td>\n",
       "      <td>-0.545618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1027</th>\n",
       "      <td>139.4</td>\n",
       "      <td>108.6</td>\n",
       "      <td>192.7</td>\n",
       "      <td>6.1</td>\n",
       "      <td>892.4</td>\n",
       "      <td>780.0</td>\n",
       "      <td>28</td>\n",
       "      <td>23.696601</td>\n",
       "      <td>-0.108017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1028</th>\n",
       "      <td>186.7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>175.6</td>\n",
       "      <td>11.3</td>\n",
       "      <td>989.6</td>\n",
       "      <td>788.9</td>\n",
       "      <td>28</td>\n",
       "      <td>32.768036</td>\n",
       "      <td>0.460269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1029</th>\n",
       "      <td>100.5</td>\n",
       "      <td>78.3</td>\n",
       "      <td>200.6</td>\n",
       "      <td>8.6</td>\n",
       "      <td>864.5</td>\n",
       "      <td>761.5</td>\n",
       "      <td>28</td>\n",
       "      <td>32.401235</td>\n",
       "      <td>-0.124507</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>956 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      BlastFurnaceSlag  FlyAsh  Water  Superplasticizer  CoarseAggregate  \\\n",
       "0                  0.0     0.0  162.0               2.5           1040.0   \n",
       "1                  0.0     0.0  162.0               2.5           1055.0   \n",
       "4                132.4     0.0  192.0               0.0            978.4   \n",
       "5                114.0     0.0  228.0               0.0            932.0   \n",
       "7                 95.0     0.0  228.0               0.0            932.0   \n",
       "...                ...     ...    ...               ...              ...   \n",
       "1025             116.0    90.3  179.6               8.9            870.1   \n",
       "1026               0.0   115.6  196.0              10.4            817.9   \n",
       "1027             139.4   108.6  192.7               6.1            892.4   \n",
       "1028             186.7     0.0  175.6              11.3            989.6   \n",
       "1029             100.5    78.3  200.6               8.6            864.5   \n",
       "\n",
       "      FineAggregate  Age  ConcreteCompressiveStrength  Residuals  \n",
       "0             676.0   28                    79.986111   1.631719  \n",
       "1             676.0   28                    61.887366   0.378786  \n",
       "4             825.5  360                    44.296075   1.038994  \n",
       "5             670.0   90                    47.029847   0.857170  \n",
       "7             594.0   28                    36.447770  -0.499601  \n",
       "...             ...  ...                          ...        ...  \n",
       "1025          768.3   28                    44.284354   0.610542  \n",
       "1026          813.4   28                    31.178794  -0.545618  \n",
       "1027          780.0   28                    23.696601  -0.108017  \n",
       "1028          788.9   28                    32.768036   0.460269  \n",
       "1029          761.5   28                    32.401235  -0.124507  \n",
       "\n",
       "[956 rows x 9 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Collective Feature Efficiency\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "\n",
    "target_column_concrete = ['ConcreteCompressiveStrength']\n",
    "\n",
    "Corr_y = ConcreteData_df.corr()['ConcreteCompressiveStrength'].abs().sort_values(ascending = True)\n",
    "closepair_df = ConcreteData_df[['Cement', 'ConcreteCompressiveStrength']]\n",
    "\n",
    "closepair_df_y = closepair_df[target_column_concrete]\n",
    "closepair_df_X = closepair_df.drop(target_column_concrete, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "standardized_residuals = influence_linear.resid_studentized_internal\n",
    "\n",
    "ConcreteData_df = ConcreteData_df.drop(['Cement'], axis = 1)\n",
    "ConcreteData_df['Residuals'] = standardized_residuals\n",
    "\n",
    "ConcreteData_df = ConcreteData_df[abs(ConcreteData_df.Residuals)  > 0.1]\n",
    "ConcreteData_df\n",
    "\n",
    "# linear_reg_concrete = LinearRegression().fit(closepair_df_X, closepair_df_y)\n",
    "# pred_linear_concrete = linear_reg.predict(closepair_df_X)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.20618256026520188"
      ]
     },
     "execution_count": 171,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Distance from Linear Function\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "target_column_concrete = ['ConcreteCompressiveStrength']\n",
    "\n",
    "closepair_df_y = ConcreteData_df[target_column_concrete]\n",
    "closepair_df_X = ConcreteData_df.drop(target_column_concrete, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "std_res = influence_linear.resid_studentized_internal\n",
    "std_res_abs = [abs(x) for x in std_res]\n",
    "\n",
    "ratio = 1\n",
    "for i in std_res_abs:\n",
    "    ratio = ratio - float(i)/len(std_res_abs)  \n",
    "    \n",
    "ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1030, 9)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.7198972540487869"
      ]
     },
     "execution_count": 223,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Input Distribution\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "import math\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "ConcreteData_df_sort = ConcreteData_df.sort_values('ConcreteCompressiveStrength')\n",
    "print(ConcreteData_df_sort.shape)\n",
    "\n",
    "target_column_concrete = ['ConcreteCompressiveStrength']\n",
    "\n",
    "ConcreteData_df_sort_X = ConcreteData_df_sort.drop(target_column_concrete, axis = 1)\n",
    "scaler = MinMaxScaler() \n",
    "scaled_values = scaler.fit_transform(ConcreteData_df_sort_X) \n",
    "ConcreteData_df_sort_X.loc[:,:] = scaled_values\n",
    "\n",
    "euclidean_pair_consec = np.linalg.norm(ConcreteData_df_sort_X.diff(axis=0).drop(0), axis=1)\n",
    "euclidean_pair_consec = [x for x in euclidean_pair_consec if math.isnan(x) == False]\n",
    "\n",
    "len(euclidean_pair_consec)\n",
    "input_dist = sum(euclidean_pair_consec)/float(ConcreteData_df_sort_X.shape[0])\n",
    "input_dist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "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>ConcreteCompressiveStrength</th>\n",
       "      <th>Residuals</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>79.986111</td>\n",
       "      <td>2.131597</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>61.887366</td>\n",
       "      <td>1.220355</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>44.296075</td>\n",
       "      <td>0.334662</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>47.029847</td>\n",
       "      <td>0.472303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>45.854291</td>\n",
       "      <td>0.413115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1016</th>\n",
       "      <td>15.091251</td>\n",
       "      <td>-0.779398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1020</th>\n",
       "      <td>42.140084</td>\n",
       "      <td>0.226111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1025</th>\n",
       "      <td>44.284354</td>\n",
       "      <td>0.495907</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1027</th>\n",
       "      <td>23.696601</td>\n",
       "      <td>-0.507717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1029</th>\n",
       "      <td>32.401235</td>\n",
       "      <td>-0.123734</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>679 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      ConcreteCompressiveStrength  Residuals\n",
       "0                       79.986111   2.131597\n",
       "1                       61.887366   1.220355\n",
       "4                       44.296075   0.334662\n",
       "5                       47.029847   0.472303\n",
       "8                       45.854291   0.413115\n",
       "...                           ...        ...\n",
       "1016                    15.091251  -0.779398\n",
       "1020                    42.140084   0.226111\n",
       "1025                    44.284354   0.495907\n",
       "1027                    23.696601  -0.507717\n",
       "1029                    32.401235  -0.123734\n",
       "\n",
       "[679 rows x 2 columns]"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# ConcreteData_df = ConcreteData_df.drop(['Residuals'], axis = 1)\n",
    "\n",
    "target_column_concrete = ['ConcreteCompressiveStrength']\n",
    "\n",
    "Corr_y = ConcreteData_df.corr()['ConcreteCompressiveStrength'].abs().sort_values(ascending = True)\n",
    "Corr_y\n",
    "closepair_df = ConcreteData_df[['FlyAsh', 'ConcreteCompressiveStrength']]\n",
    "\n",
    "closepair_df_y = closepair_df[target_column_concrete]\n",
    "closepair_df_X = closepair_df.drop(target_column_concrete, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "standardized_residuals = influence_linear.resid_studentized_internal\n",
    "\n",
    "ConcreteData_df = ConcreteData_df.drop(['FlyAsh'], axis = 1)\n",
    "ConcreteData_df['Residuals'] = standardized_residuals\n",
    "\n",
    "ConcreteData_df = ConcreteData_df[abs(ConcreteData_df.Residuals)  > 0.1]\n",
    "ConcreteData_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 224,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Housing Data\n",
      "---------------------------\n",
      "(506, 14)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Index(['crim', 'zn', 'indus', 'chas', 'nox', 'rm', 'age', 'dis', 'rad', 'tax',\n",
       "       'ptratio', 'b', 'lstat', 'medv'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 224,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "################################### Housing ################################################################\n",
    "HousingData_df = pd.read_csv('UCI_regression/BostonHousing/BostonHousing.csv') \n",
    "print(\"Housing Data\")\n",
    "print(\"---------------------------\")\n",
    "print(HousingData_df.shape)\n",
    "\n",
    "target_column_housing = ['medv']\n",
    "HousingData_df_y = HousingData_df[target_column_housing]\n",
    "HousingData_df_X = HousingData_df.drop(target_column_housing, axis = 1)\n",
    "\n",
    "HousingData_df.columns\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([4.86837097, 4.71732544, 6.02606388, 3.55534806, 6.20547279,\n",
       "       6.22032961, 6.12673135, 6.08397546, 5.86043414, 6.14553124,\n",
       "       6.21940643, 6.17532214, 6.07395666, 6.14828567])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get Entropy Housing\n",
    "entropy_housing = entropy(HousingData_df)\n",
    "entropy_housing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5292293626047331"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Housing\n",
    "corr_housing =  HousingData_df.corr(method = 'spearman')['medv'].abs()\n",
    "corr_housing = corr_housing.sort_values(ascending = True)\n",
    "corr_housing\n",
    "\n",
    "sumval_housing = corr_housing.sum()/len(corr_housing)\n",
    "sumval_housing\n",
    "\n",
    "## 7.6% simple (>0.8)\n",
    "## 15.3% simple (>0.6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.053822519683534936"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Housing for the features\n",
    "corr_housing =  HousingData_df_X.corr(method = 'spearman')\n",
    "corr_housing\n",
    "\n",
    "corr_housing_lsum = np.tril(corr_housing).sum() - np.trace(corr_housing)\n",
    "housing_elems = len(corr_housing) * (len(corr_housing) - 1) / 2\n",
    "corr_housing_lavg = abs(corr_housing_lsum / housing_elems)\n",
    "corr_housing_lavg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.2708628109003177"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the L1 parameter Housing\n",
    "from sklearn import linear_model\n",
    "\n",
    "regr_housing = linear_model.LinearRegression()\n",
    "regr_housing.fit(HousingData_df_X, HousingData_df_y)\n",
    "predict_housing = regr_housing.predict(HousingData_df_X)\n",
    "\n",
    "housing_df_y_list = HousingData_df_y.to_numpy()\n",
    "mandis_housing = np.abs(housing_df_y_list - predict_housing)\n",
    "L1_housing = np.sum(mandis_housing)/len(mandis_housing)\n",
    "L1_housing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Collective Feature Efficiency\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "HousingData_df = HousingData_df.drop(['Residuals'], axis = 1)\n",
    "\n",
    "target_column_housing = ['medv']\n",
    "\n",
    "Corr_y = HousingData_df.corr()['medv'].abs().sort_values(ascending = True)\n",
    "Corr_y\n",
    "closepair_df = HousingData_df[['dis', 'medv']]\n",
    "\n",
    "closepair_df_y = closepair_df[target_column_housing]\n",
    "closepair_df_X = closepair_df.drop(target_column_housing, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "standardized_residuals = influence_linear.resid_studentized_internal\n",
    "\n",
    "HousingData_df = HousingData_df.drop(['dis'], axis = 1)\n",
    "HousingData_df['Residuals'] = standardized_residuals\n",
    "\n",
    "HousingData_df = HousingData_df[abs(HousingData_df.Residuals)  > 0.1]\n",
    "HousingData_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.29908986791567527"
      ]
     },
     "execution_count": 174,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Distance from Linear Function\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "target_column_housing = ['medv']\n",
    "\n",
    "closepair_df_y = HousingData_df[target_column_housing]\n",
    "closepair_df_X = HousingData_df.drop(target_column_housing, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "std_res = influence_linear.resid_studentized_internal\n",
    "std_res_abs = [abs(x) for x in std_res]\n",
    "\n",
    "ratio = 1\n",
    "for i in std_res_abs:\n",
    "    ratio = ratio - float(i)/len(std_res_abs)  \n",
    "    \n",
    "ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(506, 14)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.9021480317169385"
      ]
     },
     "execution_count": 225,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Input Distribution\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "import math\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "HousingData_df_sort = HousingData_df.sort_values('medv')\n",
    "print(HousingData_df_sort.shape)\n",
    "\n",
    "target_column_housing = ['medv']\n",
    "\n",
    "HousingData_df_sort_X = HousingData_df_sort.drop(target_column_housing, axis = 1)\n",
    "scaler = MinMaxScaler() \n",
    "scaled_values = scaler.fit_transform(HousingData_df_sort_X) \n",
    "HousingData_df_sort_X.loc[:,:] = scaled_values\n",
    "\n",
    "euclidean_pair_consec = np.linalg.norm(HousingData_df_sort_X.diff(axis=0).drop(0), axis=1)\n",
    "euclidean_pair_consec = [x for x in euclidean_pair_consec if math.isnan(x) == False]\n",
    "\n",
    "len(euclidean_pair_consec)\n",
    "input_dist = sum(euclidean_pair_consec)/float(HousingData_df_sort_X.shape[0])\n",
    "input_dist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Automobile Data\n",
      "---------------------------\n",
      "The shape of the Input data is:  (392, 8)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Index(['mpg', 'cylinders', 'displacement', 'horsepower', 'weight',\n",
       "       'acceleration', 'year', 'origin'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 226,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "######################################################## Automobile ################################################################\n",
    "print(\"Automobile Data\")\n",
    "print(\"---------------------------\")\n",
    "dropcol_initial_auto = ['name']\n",
    "AutoData_df = pd.read_csv('UCI_regression/MPG/Auto.csv') ## horsepower column has correlation 0.4 :: [46 - 230] :: 30\n",
    "AutoData_df = AutoData_df.drop(dropcol_initial_auto, axis = 1)\n",
    "print(\"The shape of the Input data is: \", AutoData_df.shape)\n",
    "\n",
    "target_column_auto = ['mpg']\n",
    "\n",
    "AutoData_df_y = AutoData_df[target_column_auto]\n",
    "AutoData_df_X = AutoData_df.drop(target_column_auto, axis = 1)\n",
    "\n",
    "AutoData_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([5.91655392, 5.92424156, 5.83322921, 5.9085298 , 5.93161924,\n",
       "       5.95554068, 5.97008906, 5.85263873])"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get Entropy Auto\n",
    "entropy_auto = entropy(AutoData_df)\n",
    "entropy_auto"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7505589879419527"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Auto\n",
    "corr_auto =  AutoData_df.corr(method = 'spearman')['mpg'].abs()\n",
    "corr_auto = corr_auto.sort_values(ascending = True)\n",
    "corr_auto\n",
    "\n",
    "sumval_auto = corr_auto.sum()/len(corr_auto)\n",
    "sumval_auto\n",
    "\n",
    "## 57% simple (>0.8)\n",
    "## 57% simple (>0.6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.008094763976858784"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Auto for the features\n",
    "corr_auto =  AutoData_df_X.corr(method = 'spearman')\n",
    "corr_auto\n",
    "\n",
    "corr_auto_lsum = np.tril(corr_auto).sum() - np.trace(corr_auto)\n",
    "auto_elems = len(corr_auto) * (len(corr_auto) - 1) / 2\n",
    "corr_auto_lavg = abs(corr_auto_lsum / auto_elems)\n",
    "corr_auto_lavg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.4993098325008716"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the L1 parameter Auto\n",
    "from sklearn import linear_model\n",
    "\n",
    "regr_auto = linear_model.LinearRegression()\n",
    "regr_auto.fit(AutoData_df_X, AutoData_df_y)\n",
    "predict_auto = regr_auto.predict(AutoData_df_X)\n",
    "\n",
    "AutoData_df_y_list = AutoData_df_y.to_numpy()\n",
    "mandis_auto = np.abs(AutoData_df_y_list - predict_auto)\n",
    "L1_auto = np.sum(mandis_auto)/len(mandis_auto)\n",
    "L1_auto"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "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>mpg</th>\n",
       "      <th>Residuals</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>18.0</td>\n",
       "      <td>-0.230596</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>15.0</td>\n",
       "      <td>-0.529800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>15.0</td>\n",
       "      <td>-0.317425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>15.0</td>\n",
       "      <td>-0.103325</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>15.0</td>\n",
       "      <td>-0.317425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>381</th>\n",
       "      <td>38.0</td>\n",
       "      <td>1.517670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>385</th>\n",
       "      <td>36.0</td>\n",
       "      <td>1.842228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>386</th>\n",
       "      <td>27.0</td>\n",
       "      <td>0.123902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>388</th>\n",
       "      <td>44.0</td>\n",
       "      <td>1.204596</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>391</th>\n",
       "      <td>31.0</td>\n",
       "      <td>0.318467</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>199 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      mpg  Residuals\n",
       "0    18.0  -0.230596\n",
       "1    15.0  -0.529800\n",
       "5    15.0  -0.317425\n",
       "9    15.0  -0.103325\n",
       "10   15.0  -0.317425\n",
       "..    ...        ...\n",
       "381  38.0   1.517670\n",
       "385  36.0   1.842228\n",
       "386  27.0   0.123902\n",
       "388  44.0   1.204596\n",
       "391  31.0   0.318467\n",
       "\n",
       "[199 rows x 2 columns]"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Collective Feature Efficiency\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "# AutoData_df = AutoData_df.drop(['Residuals'], axis = 1)\n",
    "\n",
    "target_column_auto = ['mpg']\n",
    "\n",
    "Corr_y = AutoData_df.corr()['mpg'].abs().sort_values(ascending = True)\n",
    "Corr_y\n",
    "closepair_df = AutoData_df[['acceleration', 'mpg']]\n",
    "\n",
    "closepair_df_y = closepair_df[target_column_auto]\n",
    "closepair_df_X = closepair_df.drop(target_column_auto, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "standardized_residuals = influence_linear.resid_studentized_internal\n",
    "\n",
    "AutoData_df = AutoData_df.drop(['acceleration'], axis = 1)\n",
    "AutoData_df['Residuals'] = standardized_residuals\n",
    "\n",
    "AutoData_df = AutoData_df[abs(AutoData_df.Residuals)  > 0.1]\n",
    "AutoData_df\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.24043937249727487"
      ]
     },
     "execution_count": 176,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Distance from Linear Function\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "target_column_auto = ['mpg']\n",
    "\n",
    "closepair_df_y = AutoData_df[target_column_auto]\n",
    "closepair_df_X = AutoData_df.drop(target_column_auto, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "std_res = influence_linear.resid_studentized_internal\n",
    "std_res_abs = [abs(x) for x in std_res]\n",
    "\n",
    "ratio = 1\n",
    "for i in std_res_abs:\n",
    "    ratio = ratio - float(i)/len(std_res_abs)  \n",
    "    \n",
    "ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(392, 8)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.5897347838735368"
      ]
     },
     "execution_count": 227,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Input Distribution\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "import math\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "AutoData_df_sort = AutoData_df.sort_values('mpg')\n",
    "print(AutoData_df_sort.shape)\n",
    "\n",
    "target_column_auto = ['mpg']\n",
    "\n",
    "AutoData_df_sort_X = AutoData_df_sort.drop(target_column_auto, axis = 1)\n",
    "scaler = MinMaxScaler() \n",
    "scaled_values = scaler.fit_transform(AutoData_df_sort_X) \n",
    "AutoData_df_sort_X.loc[:,:] = scaled_values\n",
    "\n",
    "euclidean_pair_consec = np.linalg.norm(AutoData_df_sort_X.diff(axis=0).drop(0), axis=1)\n",
    "euclidean_pair_consec = [x for x in euclidean_pair_consec if math.isnan(x) == False]\n",
    "\n",
    "len(euclidean_pair_consec)\n",
    "input_dist = sum(euclidean_pair_consec)/float(AutoData_df_sort_X.shape[0])\n",
    "input_dist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ailerons Data\n",
      "-------------------------------------------\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(7154, 41)"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "################################################################ Ailerons ###############################################\n",
    "print(\"Ailerons Data\")\n",
    "print(\"-------------------------------------------\")\n",
    "\n",
    "\n",
    "target_ailerons = ['goal']\n",
    "colnames_ailerons = ['climbRate', 'Sgz', 'p', 'q', 'curPitch', 'curRoll', 'absRoll', 'diffClb', 'diffRollRate', 'diffDiffClb', 'SeTime1',\n",
    "            'SeTime2', 'SeTime3', 'SeTime4', 'SeTime5', 'SeTime6', 'SeTime7', 'SeTime8', 'SeTime9', 'SeTime10', 'SeTime11', 'SeTime12', \n",
    "            'SeTime13', 'SeTime14', 'diffSeTime1', 'diffSeTime2', 'diffSeTime3', 'diffSeTime4', 'diffSeTime5', 'diffSeTime6', 'diffSeTime7',\n",
    "            'diffSeTime8', 'diffSeTime9', 'diffSeTime10', 'diffSeTime11', 'diffSeTime12', 'diffSeTime13', 'diffSeTime14', 'alpha', 'Se', 'goal']\n",
    "AileronsData_train_df = pd.read_csv('UCI_regression/Ailerons/ailerons.data', header = None, names = colnames_ailerons) \n",
    "# print(AileronsData_df_train.shape)\n",
    "\n",
    "AileronsData_test_df = pd.read_csv('UCI_regression/Ailerons/ailerons.test', header = None, names = colnames_ailerons) \n",
    "# print(AileronsData_df_test.shape)\n",
    "\n",
    "AileronsData_train_df_y = AileronsData_train_df[target_ailerons]\n",
    "AileronsData_train_df_X = AileronsData_train_df.drop(target_ailerons, axis = 1)\n",
    "\n",
    "AileronsData_train_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/shrey/anaconda3/lib/python3.8/site-packages/scipy/stats/_distn_infrastructure.py:2664: RuntimeWarning: divide by zero encountered in true_divide\n",
      "  pk = 1.0*pk / np.sum(pk, axis=axis, keepdims=True)\n",
      "/Users/shrey/anaconda3/lib/python3.8/site-packages/scipy/stats/_distn_infrastructure.py:2664: RuntimeWarning: invalid value encountered in true_divide\n",
      "  pk = 1.0*pk / np.sum(pk, axis=axis, keepdims=True)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([      -inf,       -inf,       -inf,       -inf,       -inf,\n",
       "             -inf, 8.80691841,       -inf,       -inf,       -inf,\n",
       "       8.8315046 , 8.83126288, 8.83126288, 8.83121856, 8.83121697,\n",
       "       8.83117994, 8.83117394, 8.83118849, 8.83119159, 8.83106317,\n",
       "       8.83106452, 8.83094737, 8.83095125, 8.83092714,       -inf,\n",
       "              nan,       -inf, 0.        ,       -inf, 0.50040242,\n",
       "             -inf,        nan,       -inf,       -inf,       -inf,\n",
       "             -inf,       -inf, 0.69314718, 8.83117899, 8.83092714,\n",
       "       8.7747889 ])"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get Entropy Ailerons\n",
    "entropy_ailerons = entropy(AileronsData_train_df)\n",
    "entropy_ailerons"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.3217586099480845"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Ailerons\n",
    "corr_ailerons =  AileronsData_train_df.corr(method = 'spearman')['goal'].abs()\n",
    "corr_ailerons = corr_ailerons.sort_values(ascending = True)\n",
    "corr_ailerons = corr_ailerons[:-1]\n",
    "corr_ailerons\n",
    "\n",
    "sumval_ailerons = corr_ailerons.sum()/len(corr_ailerons)\n",
    "sumval_ailerons\n",
    "\n",
    "\n",
    "## 0% simple (>0.8)\n",
    "## 0% simple (>0.6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.15845264476831053"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Ailerons for the features\n",
    "AileronsData_train_df_X = AileronsData_train_df_X.drop(['diffSeTime2'], axis = 1)\n",
    "corr_ailerons =  AileronsData_train_df_X.corr(method = 'spearman')\n",
    "corr_ailerons\n",
    "\n",
    "corr_ailerons_lsum = np.tril(corr_ailerons).sum() - np.trace(corr_ailerons)\n",
    "\n",
    "ailerons_elems = len(corr_ailerons) * (len(corr_ailerons) - 1) / 2\n",
    "corr_ailerons_lavg = abs(corr_ailerons_lsum / ailerons_elems)\n",
    "corr_ailerons_lavg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.00012685091380807255"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import linear_model\n",
    "\n",
    "regr_ailerons = linear_model.LinearRegression()\n",
    "regr_ailerons.fit(AileronsData_train_df_X, AileronsData_train_df_y)\n",
    "predict_ailerons = regr_ailerons.predict(AileronsData_train_df_X)\n",
    "\n",
    "AileronsData_train_df_y_list = AileronsData_train_df_y.to_numpy()\n",
    "mandis_ailerons = np.abs(AileronsData_train_df_y_list - predict_ailerons)\n",
    "L1_ailerons = np.sum(mandis_ailerons)/len(mandis_ailerons)\n",
    "L1_ailerons\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "goal           1.0\n",
       "diffSeTime2    NaN\n",
       "diffSeTime4    NaN\n",
       "diffSeTime6    NaN\n",
       "Name: goal, dtype: float64"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Collective Feature Efficiency\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "target_column_ailerons = ['goal']\n",
    "\n",
    "AileronsData_train_df = AileronsData_train_df.drop(['Residuals'], axis = 1)\n",
    "Corr_y = AileronsData_train_df.corr()['goal'].abs().sort_values(ascending = True)\n",
    "Corr_y\n",
    "\n",
    "# closepair_df = AileronsData_train_df[['diffSeTime12', 'goal']]\n",
    "# closepair_df_y = closepair_df[target_column_ailerons]\n",
    "# closepair_df_X = closepair_df.drop(target_column_ailerons, axis = 1)\n",
    "\n",
    "# x_const = sm.add_constant(closepair_df_X)\n",
    "# linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "# influence_linear = linear_reg.get_influence()\n",
    "# standardized_residuals = influence_linear.resid_studentized_internal\n",
    "\n",
    "# AileronsData_train_df = AileronsData_train_df.drop(['diffSeTime12'], axis = 1)\n",
    "# AileronsData_train_df['Residuals'] = standardized_residuals\n",
    "\n",
    "# AileronsData_train_df = AileronsData_train_df[abs(AileronsData_train_df.Residuals)  > 0.1]\n",
    "# AileronsData_train_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.47330164942689407"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res = (795 + 586 + 17 + 26 + 62 + 0 + 22 + 0 + 23 + 0 + 29 + 0 + 30 + 0 + 37 + 0 + 171 + 375 + 358 + 284 + 223 + 151 + 110 + 37 + 20 + 16 + 3 + 6 + 1 + 3\n",
    "      + 0 + 0 + 1 + 0 + 0 + 0 + 0 + 0 + 0 + 0)/float(7154)\n",
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.2668927403644471"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Distance from Linear Function\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "import math\n",
    "\n",
    "target_column_ailerons = ['goal']\n",
    "\n",
    "closepair_df_y = AileronsData_train_df[target_column_ailerons]\n",
    "closepair_df_X = AileronsData_train_df.drop(target_column_ailerons, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "std_res = influence_linear.resid_studentized_internal\n",
    "std_res_abs = [abs(x) for x in std_res]\n",
    "# print(std_res)\n",
    "\n",
    "ratio = 1\n",
    "for i in std_res_abs:\n",
    "    if math.isnan(i) is False:\n",
    "        ratio = ratio - float(i)/len(std_res_abs)  \n",
    "    \n",
    "ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7154, 41)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.6868697527730712"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Input Distribution\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "import math\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "AileronsData_train_df_sort = AileronsData_train_df.sort_values('goal')\n",
    "print(AileronsData_train_df_sort.shape)\n",
    "\n",
    "target_column_ailerons = ['goal']\n",
    "\n",
    "AileronsData_train_df_sort_X = AileronsData_train_df_sort.drop(target_column_ailerons, axis = 1)\n",
    "scaler = MinMaxScaler() \n",
    "scaled_values = scaler.fit_transform(AileronsData_train_df_sort_X) \n",
    "AileronsData_train_df_sort_X.loc[:,:] = scaled_values\n",
    "\n",
    "euclidean_pair_consec = np.linalg.norm(AileronsData_train_df_sort_X.diff(axis=0).drop(0), axis=1)\n",
    "euclidean_pair_consec = [x for x in euclidean_pair_consec if math.isnan(x) == False]\n",
    "\n",
    "input_dist = sum(euclidean_pair_consec)/float(AileronsData_train_df_sort_X.shape[0])\n",
    "input_dist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Elevators\n",
      "---------------------------\n",
      "Elevators Data\n",
      "(8752, 19)\n",
      "(7847, 19)\n"
     ]
    },
    {
     "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>climbRate</th>\n",
       "      <th>Sgz</th>\n",
       "      <th>p</th>\n",
       "      <th>q</th>\n",
       "      <th>curRoll</th>\n",
       "      <th>absRoll</th>\n",
       "      <th>diffClb</th>\n",
       "      <th>diffRollRate</th>\n",
       "      <th>diffDiffClb</th>\n",
       "      <th>SaTime1</th>\n",
       "      <th>SaTime2</th>\n",
       "      <th>SaTime3</th>\n",
       "      <th>SaTime4</th>\n",
       "      <th>diffSaTime1</th>\n",
       "      <th>diffSaTime2</th>\n",
       "      <th>diffSaTime3</th>\n",
       "      <th>diffSaTime4</th>\n",
       "      <th>Sa</th>\n",
       "      <th>Goal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>118</td>\n",
       "      <td>-55</td>\n",
       "      <td>-0.28</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>-0.2</td>\n",
       "      <td>-11</td>\n",
       "      <td>11</td>\n",
       "      <td>0.005</td>\n",
       "      <td>-0.2</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>0.031</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>390</td>\n",
       "      <td>-45</td>\n",
       "      <td>-0.06</td>\n",
       "      <td>-0.07</td>\n",
       "      <td>-0.6</td>\n",
       "      <td>-12</td>\n",
       "      <td>11</td>\n",
       "      <td>0.010</td>\n",
       "      <td>-0.2</td>\n",
       "      <td>-0.0008</td>\n",
       "      <td>-0.0008</td>\n",
       "      <td>-0.0008</td>\n",
       "      <td>-0.0008</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0008</td>\n",
       "      <td>0.034</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>68</td>\n",
       "      <td>6</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.15</td>\n",
       "      <td>0.6</td>\n",
       "      <td>-10</td>\n",
       "      <td>-9</td>\n",
       "      <td>-0.003</td>\n",
       "      <td>-0.2</td>\n",
       "      <td>-0.0011</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>-0.0002</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>0.033</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-358</td>\n",
       "      <td>-12</td>\n",
       "      <td>-0.20</td>\n",
       "      <td>0.13</td>\n",
       "      <td>-0.3</td>\n",
       "      <td>-11</td>\n",
       "      <td>-7</td>\n",
       "      <td>0.001</td>\n",
       "      <td>-0.1</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>0.032</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-411</td>\n",
       "      <td>-19</td>\n",
       "      <td>-0.18</td>\n",
       "      <td>0.02</td>\n",
       "      <td>-0.5</td>\n",
       "      <td>-11</td>\n",
       "      <td>-3</td>\n",
       "      <td>0.002</td>\n",
       "      <td>1.2</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0010</td>\n",
       "      <td>0.030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8747</th>\n",
       "      <td>45</td>\n",
       "      <td>20</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>0.07</td>\n",
       "      <td>-0.7</td>\n",
       "      <td>-9</td>\n",
       "      <td>-2</td>\n",
       "      <td>0.001</td>\n",
       "      <td>-1.5</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>0.017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8748</th>\n",
       "      <td>55</td>\n",
       "      <td>23</td>\n",
       "      <td>0.19</td>\n",
       "      <td>0.02</td>\n",
       "      <td>-0.3</td>\n",
       "      <td>-8</td>\n",
       "      <td>1</td>\n",
       "      <td>0.004</td>\n",
       "      <td>-0.1</td>\n",
       "      <td>-0.0004</td>\n",
       "      <td>-0.0004</td>\n",
       "      <td>-0.0004</td>\n",
       "      <td>-0.0004</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0004</td>\n",
       "      <td>0.018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8749</th>\n",
       "      <td>-107</td>\n",
       "      <td>21</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-8</td>\n",
       "      <td>-3</td>\n",
       "      <td>-0.004</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>-0.0006</td>\n",
       "      <td>-0.0006</td>\n",
       "      <td>-0.0006</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0006</td>\n",
       "      <td>0.017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8750</th>\n",
       "      <td>105</td>\n",
       "      <td>19</td>\n",
       "      <td>-0.10</td>\n",
       "      <td>-0.03</td>\n",
       "      <td>-0.7</td>\n",
       "      <td>-9</td>\n",
       "      <td>5</td>\n",
       "      <td>0.002</td>\n",
       "      <td>-0.1</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>0.018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8751</th>\n",
       "      <td>129</td>\n",
       "      <td>22</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>0.03</td>\n",
       "      <td>-0.8</td>\n",
       "      <td>-9</td>\n",
       "      <td>-3</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.4</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0005</td>\n",
       "      <td>0.017</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8752 rows × 19 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      climbRate  Sgz     p     q  curRoll  absRoll  diffClb  diffRollRate  \\\n",
       "0           118  -55 -0.28 -0.08     -0.2      -11       11         0.005   \n",
       "1           390  -45 -0.06 -0.07     -0.6      -12       11         0.010   \n",
       "2            68    6  0.11  0.15      0.6      -10       -9        -0.003   \n",
       "3          -358  -12 -0.20  0.13     -0.3      -11       -7         0.001   \n",
       "4          -411  -19 -0.18  0.02     -0.5      -11       -3         0.002   \n",
       "...         ...  ...   ...   ...      ...      ...      ...           ...   \n",
       "8747         45   20 -0.02  0.07     -0.7       -9       -2         0.001   \n",
       "8748         55   23  0.19  0.02     -0.3       -8        1         0.004   \n",
       "8749       -107   21 -0.17  0.04      0.0       -8       -3        -0.004   \n",
       "8750        105   19 -0.10 -0.03     -0.7       -9        5         0.002   \n",
       "8751        129   22 -0.08  0.03     -0.8       -9       -3         0.002   \n",
       "\n",
       "      diffDiffClb  SaTime1  SaTime2  SaTime3  SaTime4  diffSaTime1  \\\n",
       "0            -0.2  -0.0010  -0.0010  -0.0010  -0.0010       0.0000   \n",
       "1            -0.2  -0.0008  -0.0008  -0.0008  -0.0008       0.0000   \n",
       "2            -0.2  -0.0011  -0.0010  -0.0010  -0.0010      -0.0002   \n",
       "3            -0.1  -0.0010  -0.0010  -0.0010  -0.0010       0.0000   \n",
       "4             1.2  -0.0010  -0.0010  -0.0010  -0.0010       0.0000   \n",
       "...           ...      ...      ...      ...      ...          ...   \n",
       "8747         -1.5  -0.0005  -0.0005  -0.0005  -0.0005       0.0000   \n",
       "8748         -0.1  -0.0004  -0.0004  -0.0004  -0.0004       0.0000   \n",
       "8749          0.0  -0.0005  -0.0006  -0.0006  -0.0006       0.0000   \n",
       "8750         -0.1  -0.0005  -0.0005  -0.0005  -0.0005       0.0000   \n",
       "8751          0.4  -0.0005  -0.0005  -0.0005  -0.0005       0.0000   \n",
       "\n",
       "      diffSaTime2  diffSaTime3  diffSaTime4      Sa   Goal  \n",
       "0             0.0          0.0          0.0 -0.0010  0.031  \n",
       "1             0.0          0.0          0.0 -0.0008  0.034  \n",
       "2             0.0          0.0          0.0 -0.0010  0.033  \n",
       "3             0.0          0.0          0.0 -0.0010  0.032  \n",
       "4             0.0          0.0          0.0 -0.0010  0.030  \n",
       "...           ...          ...          ...     ...    ...  \n",
       "8747          0.0          0.0          0.0 -0.0005  0.017  \n",
       "8748          0.0          0.0          0.0 -0.0004  0.018  \n",
       "8749          0.0          0.0          0.0 -0.0006  0.017  \n",
       "8750          0.0          0.0          0.0 -0.0005  0.018  \n",
       "8751          0.0          0.0          0.0 -0.0005  0.017  \n",
       "\n",
       "[8752 rows x 19 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "####################################### Elevators #########################################################\n",
    "print(\"Elevators\")\n",
    "print(\"---------------------------\")\n",
    "\n",
    "target_elevators = ['Goal']\n",
    "colnames_elevators = ['climbRate', 'Sgz', 'p', 'q', 'curRoll', 'absRoll', 'diffClb', 'diffRollRate', 'diffDiffClb', 'SaTime1', 'SaTime2', \n",
    "                      'SaTime3', 'SaTime4', 'diffSaTime1', 'diffSaTime2', 'diffSaTime3', 'diffSaTime4', 'Sa', 'Goal']\n",
    "Elevators_train_df = pd.read_csv('UCI_regression/Elevators/elevators.data', header = None, names = colnames_elevators)\n",
    "print(\"Elevators Data\")\n",
    "print(Elevators_train_df.shape)\n",
    "\n",
    "Elevators_test_df = pd.read_csv('UCI_regression/Elevators/elevators.test', header = None, names = colnames_elevators)\n",
    "print(Elevators_test_df.shape)\n",
    "\n",
    "Elevators_train_df.dropna()\n",
    "Elevators_train_df\n",
    "# Elevators_train_df_y = Elevators_train_df[target_elevators]\n",
    "# Elevators_train_df_X = Elevators_train_df.drop(target_elevators, axis = 1)\n",
    "\n",
    "# Elevators_train_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/shrey/anaconda3/lib/python3.8/site-packages/scipy/stats/_distn_infrastructure.py:2664: RuntimeWarning: invalid value encountered in true_divide\n",
      "  pk = 1.0*pk / np.sum(pk, axis=axis, keepdims=True)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([      -inf,       -inf,       -inf,       -inf,       -inf,\n",
       "       9.00731186,       -inf,       -inf,       -inf, 8.9453694 ,\n",
       "       8.9433354 , 8.94332638, 8.94199044,       -inf, 0.        ,\n",
       "             -inf,        nan, 8.94198058, 9.03414541])"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get Entropy Elevators\n",
    "entropy_elevators = entropy(Elevators_train_df)\n",
    "entropy_elevators"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.287034664255778"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation\n",
    "corr_elevators =  Elevators_train_df.corr(method = 'spearman')['Goal'].abs()\n",
    "corr_elevators = corr_elevators.sort_values(ascending = True)\n",
    "corr_elevators = corr_elevators[:-1]\n",
    "corr_elevators\n",
    "\n",
    "sumval_elevators = corr_elevators.sum()/len(corr_elevators)\n",
    "sumval_elevators\n",
    "\n",
    "## 0% simple (>0.8)\n",
    "## 0% simple (>0.6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.08448968016243141"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Auto for the features\n",
    "Elevators_train_df_X = Elevators_train_df_X.drop(['diffSaTime4'], axis = 1)\n",
    "corr_elevators =  Elevators_train_df_X.corr(method = 'spearman')\n",
    "corr_elevators\n",
    "\n",
    "corr_elevators_lsum = np.tril(corr_elevators).sum() - np.trace(corr_elevators)\n",
    "elevators_elems = len(corr_elevators) * (len(corr_elevators) - 1) / 2\n",
    "corr_elevators_lavg = abs(corr_elevators_lsum / elevators_elems)\n",
    "corr_elevators_lavg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.001950331259306508"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import linear_model\n",
    "\n",
    "regr_elevators = linear_model.LinearRegression()\n",
    "regr_elevators.fit(Elevators_train_df_X, Elevators_train_df_y)\n",
    "predict_elevators = regr_elevators.predict(Elevators_train_df_X)\n",
    "\n",
    "Elevators_train_df_y_list = Elevators_train_df_y.to_numpy()\n",
    "mandis_elevators = np.abs(Elevators_train_df_y_list - predict_elevators)\n",
    "L1_elevators = np.sum(mandis_elevators)/len(mandis_elevators)\n",
    "L1_elevators"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "diffSaTime2     0.008155\n",
       "diffSaTime3     0.019601\n",
       "diffRollRate    0.032380\n",
       "diffSaTime1     0.045142\n",
       "Sgz             0.054617\n",
       "diffDiffClb     0.065262\n",
       "curRoll         0.074914\n",
       "p               0.075193\n",
       "diffClb         0.096297\n",
       "climbRate       0.118730\n",
       "q               0.307703\n",
       "absRoll         0.382790\n",
       "Sa              0.632127\n",
       "SaTime4         0.632143\n",
       "SaTime3         0.638642\n",
       "SaTime2         0.638655\n",
       "SaTime1         0.651617\n",
       "Goal            1.000000\n",
       "diffSaTime4          NaN\n",
       "Name: Goal, dtype: float64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Collective Feature Efficiency\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "# Elevators_train_df = Elevators_train_df.drop(['Residuals'], axis = 1)\n",
    "\n",
    "target_column_elevators = ['Goal']\n",
    "\n",
    "Corr_y = Elevators_train_df.corr()['Goal'].abs().sort_values(ascending = True)\n",
    "Corr_y\n",
    "# closepair_df = Elevators_train_df[['diffSaTime3', 'Goal']]\n",
    "\n",
    "# closepair_df_y = closepair_df[target_column_elevators]\n",
    "# closepair_df_X = closepair_df.drop(target_column_elevators, axis = 1)\n",
    "\n",
    "# x_const = sm.add_constant(closepair_df_X)\n",
    "# linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "# influence_linear = linear_reg.get_influence()\n",
    "# standardized_residuals = influence_linear.resid_studentized_internal\n",
    "\n",
    "# Elevators_train_df = Elevators_train_df.drop(['diffSaTime3'], axis = 1)\n",
    "# Elevators_train_df['Residuals'] = standardized_residuals\n",
    "\n",
    "# Elevators_train_df = Elevators_train_df[abs(Elevators_train_df.Residuals)  > 0.1]\n",
    "# Elevators_train_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 390,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.41361974405850094"
      ]
     },
     "execution_count": 390,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res = (975 + 193 + 0 + 78 + 1 + 777 + 491 + 370 + 284 + 201 + 101 + 40 + 37 + 22 + 45 + 5)/float(8752)\n",
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.326481279298307"
      ]
     },
     "execution_count": 179,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Distance from Linear Function\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "target_column_elevators = ['Goal']\n",
    "\n",
    "closepair_df_y = Elevators_train_df[target_column_elevators]\n",
    "closepair_df_X = Elevators_train_df.drop(target_column_elevators, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "std_res = influence_linear.resid_studentized_internal\n",
    "std_res_abs = [abs(x) for x in std_res]\n",
    "\n",
    "ratio = 1\n",
    "for i in std_res_abs:\n",
    "    ratio = ratio - float(i)/len(std_res_abs)  \n",
    "    \n",
    "ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5959722476620078"
      ]
     },
     "execution_count": 230,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Input Distribution\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "import math\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "Elevators_train_df_sort = Elevators_train_df.sort_values('Goal')\n",
    "print(Elevators_train_df_sort.shape)\n",
    "\n",
    "target_column_elevators = ['Goal']\n",
    "\n",
    "Elevators_train_df_sort_X = Elevators_train_df_sort.drop(target_column_elevators, axis = 1)\n",
    "scaler = MinMaxScaler() \n",
    "scaled_values = scaler.fit_transform(Elevators_train_df_sort_X) \n",
    "Elevators_train_df_sort_X.loc[:,:] = scaled_values\n",
    "\n",
    "euclidean_pair_consec = np.linalg.norm(Elevators_train_df_sort_X.diff(axis=0).drop(0), axis=1)\n",
    "euclidean_pair_consec = [x for x in euclidean_pair_consec if math.isnan(x) == False]\n",
    "\n",
    "input_dist = sum(euclidean_pair_consec)/float(Elevators_train_df_sort_X.shape[0])\n",
    "input_dist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Abalone Data\n",
      "---------------------------\n",
      "(4177, 9)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Index(['Sex', 'Length', 'Diameter', 'Height', 'Whole_weight', 'Shucked_weight',\n",
       "       'Viscera_weight', 'Shell_weight', 'Rings'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 241,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "################################### Abalone ###########################################\n",
    "print(\"Abalone Data\")\n",
    "print(\"---------------------------\")\n",
    "\n",
    "target_var_abalone = ['Rings']\n",
    "colnames_abalone = ['Sex', 'Length', 'Diameter', 'Height', 'Whole_weight', 'Shucked_weight', 'Viscera_weight', 'Shell_weight', 'Rings']\n",
    "AbaloneData_df = pd.read_csv('UCI_regression/Abalone/abalone.data', header = None, names = colnames_abalone)\n",
    "\n",
    "gender = {'M': 1,'F': 2, 'I': 3} \n",
    "AbaloneData_df.Sex = [gender[item] for item in AbaloneData_df.Sex] \n",
    "\n",
    "print(AbaloneData_df.shape)\n",
    "\n",
    "target_column_abalone = ['Rings']\n",
    "\n",
    "AbaloneData_df_y = AbaloneData_df[target_column_abalone]\n",
    "AbaloneData_df_X = AbaloneData_df.drop(target_column_abalone, axis = 1)\n",
    "\n",
    "AbaloneData_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([8.24464687, 8.30881122, 8.30506499, 8.29346748, 8.15203316,\n",
       "       8.13996549, 8.1441803 , 8.15924304, 8.28746906])"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get Entropy Abalone\n",
    "entropy_abalone = entropy(AbaloneData_df)\n",
    "entropy_abalone"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.641097950716637"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Abalone\n",
    "corr_abalone =  AbaloneData_df.corr(method = 'spearman')['Rings'].abs()\n",
    "corr_abalone = corr_abalone.sort_values(ascending = True)\n",
    "corr_abalone\n",
    "\n",
    "sumval_abalone = corr_abalone.sum()/len(corr_abalone)\n",
    "sumval_abalone\n",
    "\n",
    "\n",
    "## 0% simple (>0.8)\n",
    "## 75% simple (>0.6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.107982683781241"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Abalone for the features\n",
    "corr_abalone =  AbaloneData_df_X.corr(method = 'spearman')\n",
    "corr_abalone\n",
    "\n",
    "corr_abalone_lsum = np.tril(corr_abalone).sum() - np.trace(corr_abalone)\n",
    "abalone_elems = len(corr_abalone) * (len(corr_abalone) - 1) / 2\n",
    "corr_abalone_lavg = abs(corr_abalone_lsum / elevators_elems)\n",
    "corr_abalone_lavg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.5843987714248777"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import linear_model\n",
    "\n",
    "regr_abalone = linear_model.LinearRegression()\n",
    "regr_abalone.fit(AbaloneData_df_X, AbaloneData_df_y)\n",
    "predict_abalone = regr_abalone.predict(AbaloneData_df_X)\n",
    "\n",
    "AbaloneData_df_y_list = AbaloneData_df_y.to_numpy()\n",
    "mandis_abalone = np.abs(AbaloneData_df_y_list - predict_abalone)\n",
    "L1_abalone = np.sum(mandis_abalone)/len(mandis_abalone)\n",
    "L1_abalone"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "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>Rings</th>\n",
       "      <th>Residuals</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>15</td>\n",
       "      <td>1.044303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>7</td>\n",
       "      <td>-1.323057</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9</td>\n",
       "      <td>-0.274760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10</td>\n",
       "      <td>-0.435297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>20</td>\n",
       "      <td>2.979654</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4166</th>\n",
       "      <td>10</td>\n",
       "      <td>0.477523</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4168</th>\n",
       "      <td>8</td>\n",
       "      <td>-0.570616</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4171</th>\n",
       "      <td>8</td>\n",
       "      <td>-1.027137</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4174</th>\n",
       "      <td>9</td>\n",
       "      <td>-0.731217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4176</th>\n",
       "      <td>12</td>\n",
       "      <td>0.156543</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2872 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Rings  Residuals\n",
       "0        15   1.044303\n",
       "1         7  -1.323057\n",
       "2         9  -0.274760\n",
       "3        10  -0.435297\n",
       "6        20   2.979654\n",
       "...     ...        ...\n",
       "4166     10   0.477523\n",
       "4168      8  -0.570616\n",
       "4171      8  -1.027137\n",
       "4174      9  -0.731217\n",
       "4176     12   0.156543\n",
       "\n",
       "[2872 rows x 2 columns]"
      ]
     },
     "execution_count": 134,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Collective Feature Efficiency\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "# AbaloneData_df = AbaloneData_df.drop(['Residuals'], axis = 1)\n",
    "\n",
    "target_column_abalone = ['Rings']\n",
    "\n",
    "Corr_y = AbaloneData_df.corr()['Rings'].abs().sort_values(ascending = True)\n",
    "Corr_y\n",
    "closepair_df = AbaloneData_df[['Sex', 'Rings']]\n",
    "\n",
    "closepair_df_y = closepair_df[target_column_abalone]\n",
    "closepair_df_X = closepair_df.drop(target_column_abalone, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "standardized_residuals = influence_linear.resid_studentized_internal\n",
    "\n",
    "AbaloneData_df = AbaloneData_df.drop(['Sex'], axis = 1)\n",
    "AbaloneData_df['Residuals'] = standardized_residuals\n",
    "\n",
    "AbaloneData_df = AbaloneData_df[abs(AbaloneData_df.Residuals)  > 0.1]\n",
    "AbaloneData_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.27840395217504593"
      ]
     },
     "execution_count": 181,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Distance from Linear Function\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "target_column_abalone = ['Rings']\n",
    "\n",
    "closepair_df_y = AbaloneData_df[target_column_abalone]\n",
    "closepair_df_X = AbaloneData_df.drop(target_column_abalone, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "std_res = influence_linear.resid_studentized_internal\n",
    "std_res_abs = [abs(x) for x in std_res]\n",
    "\n",
    "ratio = 1\n",
    "for i in std_res_abs:\n",
    "    ratio = ratio - float(i)/len(std_res_abs)  \n",
    "    \n",
    "ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(4177, 9)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.5193337315047654"
      ]
     },
     "execution_count": 242,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Input Distribution\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "import math\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "AbaloneData_df_sort = AbaloneData_df.sort_values('Rings')\n",
    "print(AbaloneData_df_sort.shape)\n",
    "\n",
    "target_column_abalone = ['Rings']\n",
    "\n",
    "AbaloneData_df_sort_X = AbaloneData_df_sort.drop(target_column_abalone, axis = 1)\n",
    "scaler = MinMaxScaler() \n",
    "scaled_values = scaler.fit_transform(AbaloneData_df_sort_X) \n",
    "AbaloneData_df_sort_X.loc[:,:] = scaled_values\n",
    "\n",
    "euclidean_pair_consec = np.linalg.norm(AbaloneData_df_sort_X.diff(axis=0).drop(0), axis=1)\n",
    "euclidean_pair_consec = [x for x in euclidean_pair_consec if math.isnan(x) == False]\n",
    "\n",
    "input_dist = sum(euclidean_pair_consec)/float(AbaloneData_df_sort_X.shape[0])\n",
    "input_dist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Kinematics Data\n",
      "---------------------------\n",
      "Kinematics Data\n",
      "(8192, 9)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Index(['theta1', 'theta2', 'theta3', 'theta4', 'theta5', 'theta6', 'theta7',\n",
       "       'theta8', 'y'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 243,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "################################################## Kinematics ######################################################################\n",
    "print(\"Kinematics Data\")\n",
    "print(\"---------------------------\")\n",
    "\n",
    "target_var_Kinematics = ['y']\n",
    "colnames_Kinematics = ['theta1', 'theta2', 'theta3', 'theta4', 'theta5', 'theta6', 'theta7', 'theta8', 'y']\n",
    "KinematicsData_df = pd.read_csv('UCI_regression/Kinematics/kin8nm.data', header = None,  names = colnames_Kinematics)\n",
    "print(\"Kinematics Data\")\n",
    "print(KinematicsData_df.shape)\n",
    "\n",
    "\n",
    "target_column_kinematics = ['y']\n",
    "\n",
    "KinematicsData_df_y = KinematicsData_df[target_column_kinematics]\n",
    "KinematicsData_df_X = KinematicsData_df.drop(target_column_kinematics, axis = 1)\n",
    "\n",
    "KinematicsData_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([      -inf,       -inf,       -inf,       -inf,       -inf,\n",
       "             -inf,       -inf,       -inf, 8.93905988])"
      ]
     },
     "execution_count": 141,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get Entropy Kinematics\n",
    "entropy_kinematics = entropy(KinematicsData_df)\n",
    "entropy_kinematics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.2675720416916285"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Kinematics\n",
    "corr_kinematics =  KinematicsData_df.corr(method = 'spearman')['y'].abs()\n",
    "corr_kinematics = corr_kinematics.sort_values(ascending = True)\n",
    "corr_kinematics\n",
    "\n",
    "sumval_kinematics = corr_kinematics.sum()/len(corr_kinematics)\n",
    "sumval_kinematics\n",
    "\n",
    "\n",
    "## 0% simple (>0.8)\n",
    "## 0% simple (>0.6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0007845271852085897"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Kinematics for the features\n",
    "corr_kinematics =  KinematicsData_df_X.corr(method = 'spearman')\n",
    "corr_kinematics\n",
    "\n",
    "corr_kinematics_lsum = np.tril(corr_kinematics).sum() - np.trace(corr_kinematics)\n",
    "kinematics_elems = len(corr_kinematics) * (len(corr_kinematics) - 1) / 2\n",
    "corr_kinematics_lavg = abs(corr_kinematics_lsum / kinematics_elems)\n",
    "corr_kinematics_lavg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8078654380177874"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import linear_model\n",
    "\n",
    "regr_kinematics = linear_model.LinearRegression()\n",
    "regr_kinematics.fit(KinematicsData_df_X, KinematicsData_df_y)\n",
    "predict_kinematics = regr_kinematics.predict(KinematicsData_df_X)\n",
    "\n",
    "KinematicsData_df_y_list = KinematicsData_df_y.to_numpy()\n",
    "mandis_kinematics = np.abs(KinematicsData_df_y_list - predict_kinematics)\n",
    "L1_kinematics = np.sum(mandis_abalone)/len(mandis_kinematics)\n",
    "L1_kinematics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "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</th>\n",
       "      <th>Residuals</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.536524</td>\n",
       "      <td>-0.541909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.308014</td>\n",
       "      <td>-1.357742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.518900</td>\n",
       "      <td>-0.779425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.494151</td>\n",
       "      <td>-0.621121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.470218</td>\n",
       "      <td>-0.783857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8185</th>\n",
       "      <td>0.574224</td>\n",
       "      <td>-0.428805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8186</th>\n",
       "      <td>1.020879</td>\n",
       "      <td>0.900994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8187</th>\n",
       "      <td>0.486093</td>\n",
       "      <td>-0.767208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8190</th>\n",
       "      <td>0.887959</td>\n",
       "      <td>0.559823</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8191</th>\n",
       "      <td>0.496853</td>\n",
       "      <td>-0.756964</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5743 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             y  Residuals\n",
       "0     0.536524  -0.541909\n",
       "1     0.308014  -1.357742\n",
       "2     0.518900  -0.779425\n",
       "3     0.494151  -0.621121\n",
       "4     0.470218  -0.783857\n",
       "...        ...        ...\n",
       "8185  0.574224  -0.428805\n",
       "8186  1.020879   0.900994\n",
       "8187  0.486093  -0.767208\n",
       "8190  0.887959   0.559823\n",
       "8191  0.496853  -0.756964\n",
       "\n",
       "[5743 rows x 2 columns]"
      ]
     },
     "execution_count": 168,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Collective Feature Efficiency\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "# KinematicsData_df = KinematicsData_df.drop(['Residuals'], axis = 1)\n",
    "\n",
    "target_column_kinematics = ['y']\n",
    "\n",
    "Corr_y = KinematicsData_df.corr()['y'].abs().sort_values(ascending = True)\n",
    "Corr_y\n",
    "closepair_df = KinematicsData_df[['theta2', 'y']]\n",
    "\n",
    "closepair_df_y = closepair_df[target_column_kinematics]\n",
    "closepair_df_X = closepair_df.drop(target_column_kinematics, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "standardized_residuals = influence_linear.resid_studentized_internal\n",
    "\n",
    "KinematicsData_df = KinematicsData_df.drop(['theta2'], axis = 1)\n",
    "KinematicsData_df['Residuals'] = standardized_residuals\n",
    "\n",
    "KinematicsData_df = KinematicsData_df[abs(KinematicsData_df.Residuals)  > 0.1]\n",
    "KinematicsData_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.1968200628698036"
      ]
     },
     "execution_count": 184,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Distance from Linear Function\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "target_column_kinematics = ['y']\n",
    "\n",
    "closepair_df_y = KinematicsData_df[target_column_kinematics]\n",
    "closepair_df_X = KinematicsData_df.drop(target_column_kinematics, axis = 1)\n",
    "\n",
    "\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "std_res = influence_linear.resid_studentized_internal\n",
    "std_res_abs = [abs(x) for x in std_res]\n",
    "\n",
    "ratio = 1\n",
    "for i in std_res_abs:\n",
    "    ratio = ratio - float(i)/len(std_res_abs)  \n",
    "    \n",
    "ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8192, 9)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "1.0855204698363523"
      ]
     },
     "execution_count": 244,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Input Distribution\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "import math\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "KinematicsData_df_sort = KinematicsData_df.sort_values('y')\n",
    "print(KinematicsData_df_sort.shape)\n",
    "\n",
    "target_column_kinematics = ['y']\n",
    "\n",
    "KinematicsData_df_sort_X = KinematicsData_df_sort.drop(target_column_kinematics, axis = 1)\n",
    "scaler = MinMaxScaler() \n",
    "scaled_values = scaler.fit_transform(KinematicsData_df_sort_X) \n",
    "KinematicsData_df_sort_X.loc[:,:] = scaled_values\n",
    "\n",
    "euclidean_pair_consec = np.linalg.norm(KinematicsData_df_sort_X.diff(axis=0).drop(0), axis=1)\n",
    "euclidean_pair_consec = [x for x in euclidean_pair_consec if math.isnan(x) == False]\n",
    "\n",
    "input_dist = sum(euclidean_pair_consec)/float(KinematicsData_df_sort_X.shape[0])\n",
    "input_dist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Computer Activity Data\n",
      "---------------------------\n",
      "Computer Activity Data\n",
      "(8192, 22)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Index(['lread', 'lwrite', 'scall', 'sread', 'swrite', 'fork', 'exec', 'rchar',\n",
       "       'wchar', 'pgout', 'ppgout', 'pgfree', 'pgscan', 'atch', 'pgin', 'ppgin',\n",
       "       'pflt', 'vflt', 'runqsz', 'freemem', 'freeswap', 'usr'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 245,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "################################### Computer Activity ###########################################\n",
    "print(\"Computer Activity Data\")\n",
    "print(\"---------------------------\")\n",
    "\n",
    "colnames_CompAct = ['lread', 'lwrite', 'scall', 'sread', 'swrite', 'fork', 'exec', 'rchar', 'wchar', 'pgout', 'ppgout', \n",
    "                    'pgfree', 'pgscan', 'atch', 'pgin', 'ppgin', 'pflt', 'vflt', 'runqsz', 'freemem', 'freeswap', 'usr' ]\n",
    "CompActData_df = pd.read_csv('UCI_regression/ComputerActivity/cpu_act.data', header = None, names = colnames_CompAct)\n",
    "print(\"Computer Activity Data\")\n",
    "print(CompActData_df.shape)\n",
    "\n",
    "target_column_compact = ['usr']\n",
    "\n",
    "CompActData_df_y = CompActData_df[target_column_compact]\n",
    "CompActData_df_X = CompActData_df.drop(target_column_compact, axis = 1)\n",
    "\n",
    "CompActData_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([7.91639895, 7.6747772 , 8.75804375, 8.6847287 , 8.66940032,\n",
       "       8.37481187, 8.11147714, 8.4466664 , 8.36579607, 7.57650099,\n",
       "       7.41326948, 7.28712951, 6.90685981, 7.00746933, 8.10300314,\n",
       "       8.03916635, 8.55754906, 8.56220271, 6.09652214, 8.25778507,\n",
       "       8.94376416, 8.96897518])"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get Entropy Computer Activity\n",
    "entropy_compact = entropy(CompActData_df)\n",
    "entropy_compact"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5646829639745752"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Computer Activity\n",
    "corr_compact =  CompActData_df.corr(method = 'spearman')['usr'].abs()\n",
    "corr_compact = corr_compact.sort_values(ascending = True)\n",
    "\n",
    "sumval_compact = corr_compact.sum()/len(corr_compact)\n",
    "sumval_compact\n",
    "\n",
    "## 4.7% simple (>0.8)\n",
    "## 38% simple (>0.6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.2617228838728001"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Computer Activity for the features\n",
    "corr_compact =  CompActData_df_X.corr(method = 'spearman')\n",
    "corr_compact\n",
    "\n",
    "corr_compact_lsum = np.tril(corr_compact).sum() - np.trace(corr_compact)\n",
    "compact_elems = len(corr_compact) * (len(corr_compact) - 1) / 2\n",
    "corr_compact_lavg = abs(corr_compact_lsum / compact_elems)\n",
    "corr_compact_lavg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "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>usr</th>\n",
       "      <th>Residuals</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>92</td>\n",
       "      <td>0.307268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>93</td>\n",
       "      <td>0.355359</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>92</td>\n",
       "      <td>0.309355</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0</td>\n",
       "      <td>-2.619889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>94</td>\n",
       "      <td>0.539232</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8179</th>\n",
       "      <td>98</td>\n",
       "      <td>0.512341</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8180</th>\n",
       "      <td>92</td>\n",
       "      <td>0.298922</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8184</th>\n",
       "      <td>96</td>\n",
       "      <td>0.443287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8185</th>\n",
       "      <td>93</td>\n",
       "      <td>0.353272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8189</th>\n",
       "      <td>92</td>\n",
       "      <td>0.303095</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2921 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      usr  Residuals\n",
       "5      92   0.307268\n",
       "11     93   0.355359\n",
       "13     92   0.309355\n",
       "14      0  -2.619889\n",
       "15     94   0.539232\n",
       "...   ...        ...\n",
       "8179   98   0.512341\n",
       "8180   92   0.298922\n",
       "8184   96   0.443287\n",
       "8185   93   0.353272\n",
       "8189   92   0.303095\n",
       "\n",
       "[2921 rows x 2 columns]"
      ]
     },
     "execution_count": 308,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Collective Feature Efficiency\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "# CompActData_df = CompActData_df.drop(['Residuals'], axis = 1)\n",
    "\n",
    "target_column_compact = ['usr']\n",
    "\n",
    "Corr_y = CompActData_df.corr()['usr'].abs().sort_values(ascending = True)\n",
    "Corr_y\n",
    "closepair_df = CompActData_df[['lread', 'usr']]\n",
    "\n",
    "closepair_df_y = closepair_df[target_column_compact]\n",
    "closepair_df_X = closepair_df.drop(target_column_compact, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "standardized_residuals = influence_linear.resid_studentized_internal\n",
    "\n",
    "CompActData_df = CompActData_df.drop(['lread'], axis = 1)\n",
    "CompActData_df['Residuals'] = standardized_residuals\n",
    "\n",
    "CompActData_df = CompActData_df[abs(CompActData_df.Residuals)  > 0.1]\n",
    "CompActData_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.36813257071016225"
      ]
     },
     "execution_count": 186,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Distance from Linear Function\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "\n",
    "target_column_compact = ['usr']\n",
    "\n",
    "closepair_df_y = CompActData_df[target_column_compact]\n",
    "closepair_df_X = CompActData_df.drop(target_column_compact, axis = 1)\n",
    "\n",
    "x_const = sm.add_constant(closepair_df_X)\n",
    "linear_reg = sm.OLS(closepair_df_y, x_const).fit() \n",
    "influence_linear = linear_reg.get_influence()\n",
    "std_res = influence_linear.resid_studentized_internal\n",
    "std_res_abs = [abs(x) for x in std_res]\n",
    "\n",
    "ratio = 1\n",
    "for i in std_res_abs:\n",
    "    ratio = ratio - float(i)/len(std_res_abs)  \n",
    "    \n",
    "ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8192, 22)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.5863011680308041"
      ]
     },
     "execution_count": 238,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Input Distribution\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import statsmodels.api as sm\n",
    "import math\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "CompActData_df_sort = CompActData_df.sort_values('usr')\n",
    "print(CompActData_df_sort.shape)\n",
    "\n",
    "target_column_compact = ['usr']\n",
    "\n",
    "CompActData_df_sort = CompActData_df.drop(target_column_compact, axis = 1)\n",
    "scaler = MinMaxScaler() \n",
    "scaled_values = scaler.fit_transform(CompActData_df_sort) \n",
    "CompActData_df_sort.loc[:,:] = scaled_values\n",
    "\n",
    "euclidean_pair_consec = np.linalg.norm(CompActData_df_sort.diff(axis=0).drop(0), axis=1)\n",
    "euclidean_pair_consec = [x for x in euclidean_pair_consec if math.isnan(x) == False]\n",
    "\n",
    "input_dist = sum(euclidean_pair_consec)/float(CompActData_df_sort.shape[0])\n",
    "input_dist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.642822265625"
      ]
     },
     "execution_count": 309,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ans = (940 + 755 + 1044 + 553 + 373 + 289 + 198 + 303 + 532 + 75 + 48 + 51 + 20 + 14 + 32 + 3 + 5 + 14 + 6 + 9 +2)/float(8192)\n",
    "ans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6.047919131695916"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import linear_model\n",
    "\n",
    "regr_compact = linear_model.LinearRegression()\n",
    "regr_compact.fit(CompActData_df_X, CompActData_df_y)\n",
    "predict_compact = regr_compact.predict(CompActData_df_X)\n",
    "\n",
    "CompActData_df_y_list = CompActData_df_y.to_numpy()\n",
    "mandis_compact = np.abs(CompActData_df_y_list - predict_compact)\n",
    "L1_compact = np.sum(mandis_compact)/len(mandis_compact)\n",
    "L1_compact"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "dict_corr_80 = {'Concrete': 0, 'Housing': 7.6, 'Automobile': 57, 'Ailerons': 0, 'Elevators': 0, 'Abalone': 0, 'Kinematics': 0, 'Computer Activity': 4.7}\n",
    "dict_corr_60 = {'Concrete': 0, 'Housing': 15.3, 'Automobile': 57, 'Ailerons': 0, 'Elevators': 0, 'Abalone': 75, 'Kinematics': 0, 'Computer Activity': 38}\n",
    "\n",
    "width = 0.5\n",
    "\n",
    "plt.figure(figsize=(12,12))\n",
    "\n",
    "plt.bar(dict_corr_80.keys(), dict_corr_80.values(), width, color='navy')\n",
    "plt.xticks(rotation=80, fontsize = 20)\n",
    "plt.yticks(np.arange(0, 120, 20), fontsize = 20)\n",
    "\n",
    "plt.xlabel(\"Datasets\", fontsize = 15)\n",
    "plt.ylabel(\"Percentage Simplicity\", fontsize = 15)\n",
    "\n",
    "plt.savefig(\"Img_Percentage_Simplicity_80\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 864x864 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,12))\n",
    "\n",
    "width = 0.5\n",
    "\n",
    "plt.figure(figsize=(12,12))\n",
    "\n",
    "plt.bar(dict_corr_60.keys(), dict_corr_60.values(), width, color='maroon')\n",
    "plt.xticks(rotation=80, fontsize = 20)\n",
    "plt.yticks(np.arange(0, 120, 20), fontsize = 20)\n",
    "\n",
    "plt.xlabel(\"Datasets\", fontsize = 15)\n",
    "plt.ylabel(\"Percentage Simplicity\", fontsize = 15)\n",
    "\n",
    "plt.savefig(\"Img_Percentage_Simplicity_60\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "dict_L1 = {'Concrete': 8.2148, 'Housing': 3.2708, 'Automobile': 2.4993, 'Ailerons': 0.0001, 'Elevators': 0.0019, 'Abalone': 1.5843, 'Kinematics': 0.8078, 'Computer Activity': 6.0479}\n",
    "\n",
    "width = 0.5\n",
    "\n",
    "plt.figure(figsize=(12,12))\n",
    "\n",
    "\n",
    "plt.bar(dict_L1.keys(), dict_L1.values(), width, color='skyblue')\n",
    "plt.xticks(rotation=80, fontsize = 20)\n",
    "# plt.yticks(np.arange(0, 120, 20), fontsize = 20)\n",
    "\n",
    "plt.xlabel(\"Datasets\", fontsize = 15)\n",
    "plt.ylabel(\"L1 value\", fontsize = 15)\n",
    "\n",
    "plt.savefig(\"Img_L1_Value\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "dict_corr_avg = {'Concrete': 0.3703, 'Housing': 0.5292, 'Automobile': 0.7505, 'Ailerons': 0.3217, 'Elevators': 0.2870, 'Abalone': 0.6410, 'Kinematics': 0.2675, 'Computer Activity': 0.5646 }\n",
    "\n",
    "width = 0.5\n",
    "\n",
    "plt.figure(figsize=(12,12))\n",
    "\n",
    "\n",
    "plt.bar(dict_corr_avg.keys(), dict_corr_avg.values(), width, color='goldenrod')\n",
    "plt.xticks(rotation=80, fontsize = 20)\n",
    "# plt.yticks(np.arange(0, 120, 20), fontsize = 20)\n",
    "\n",
    "plt.xlabel(\"Datasets\", fontsize = 15)\n",
    "plt.ylabel(\"Average Correaltion\", fontsize = 15)\n",
    "\n",
    "plt.savefig(\"Img_Avg_Correlation\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAvkAAAMMCAYAAADJlLhmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzde7xtZV0v/s+XDYiiIiIoAgoqlmiauhMveOHkBSzDSyadDDONIDHLLNFTXo790izPryyU8FJaeTySYVSgoqbmNTanvICigBcIFfB+CRD2c/4YY8FksfZmrrXXmnPtZ7/fr9d4rTnHbX3nmLfPfMYzxqjWWgAAgH7sNO8CAACA1SXkAwBAZ4R8AADojJAPAACdEfIBAKAzO8+7gLV2u9vdrh144IHzLgMAgI6dc845V7TW9p53HQu6D/kHHnhgNm3aNO8yAADoWFV9ad41TNJdBwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRm53kXAADA+vLSqnmXsGIvbm3eJawLWvIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADozMxDflUdUVXnV9UFVXXiEtN/oao+OQ4fqar7TEz7YlV9qqr+o6o2zbZyAADYPuw8y39WVRuSnJTkUUkuSXJ2VZ3eWjtvYrYvJHl4a+2bVXVkklOSHDox/fDW2hUzKxoAALYzs27Jf0CSC1prF7XWrk7y1iRHTc7QWvtIa+2b492PJdl/xjUCAMB2bdYhf78kF0/cv2QctyXPSHLmxP2W5N1VdU5VHbulharq2KraVFWbLr/88m0qGAAAtjcz7a6TpJYY15acserwDCH/sInRD2mtXVpV+yQ5q6o+21r74I1W2NopGbr5ZOPGjUuuHwAAejXrlvxLkhwwcX//JJcunqmq7p3k9UmOaq19fWF8a+3S8e9lSU7L0P0HAACYMOuQf3aSg6vqoKraNcnRSU6fnKGq7pTk75P8YmvtcxPjd6+qWy3cTvLoJJ+eWeUAALCdmGl3ndbaNVV1QpJ3JdmQ5I2ttXOr6rhx+slJXpRkrySvqaokuaa1tjHJ7ZOcNo7bOclbWmvvnGX9AACwPZh1n/y01s5IcsaicSdP3H5mkmcusdxFSe6zeDwAAHBDrngLAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOjPzkF9VR1TV+VV1QVWduMT0X6iqT47DR6rqPtMuCwAAzDjkV9WGJCclOTLJIUl+vqoOWTTbF5I8vLV27yQvS3LKMpYFAIAd3qxb8h+Q5ILW2kWttauTvDXJUZMztNY+0lr75nj3Y0n2n3ZZAABg9iF/vyQXT9y/ZBy3Jc9IcuZyl62qY6tqU1Vtuvzyy7ehXAAA2P7MOuTXEuPakjNWHZ4h5D9/ucu21k5prW1srW3ce++9V1QoAABsr3ae8f+7JMkBE/f3T3Lp4pmq6t5JXp/kyNba15ezLAAA7Ohm3ZJ/dpKDq+qgqto1ydFJTp+coarulOTvk/xia+1zy1kWAACYcUt+a+2aqjohybuSbEjyxtbauVV13Dj95CQvSrJXktdUVZJcM3a9WXLZWdYPAADbg1l310lr7YwkZywad/LE7Wcmeea0ywIAADfkircAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgMztPO2NVbUzyxCT7J9lt0eTWWnvKahYGAACszFQhv6qOT/LnSb6e5PNJrl7LogAAgJWbtiX/eUn+MslxrbVr1rAeAABgG03bJ3+fJP9bwAcAgPVv2pB/ZpJD17IQAABgdUzbXeekJKdU1S5JzkryrcUztNbOW83CAACAlZk25P/L+PfFSV60aFolaUk2rFZRAADAyk0b8g9f0yoAAIBVM1XIb619YK0LAQAAVsfUF8NKkqo6NMlhSW6b5BtJPtRa+/haFAYAAKzMtBfD2j3JqUmOSHJNhoti7ZVkQ1W9M8mTW2s/WLMqAQCAqU17Cs1XJnlQkqck2a21tm+S3ZIcPY7/w7UpDwAAWK5pQ/6Tkjy/tXZqa21zkrTWNrfWTk1yYpInr1WBAADA8kwb8vdIcvEWpl2c5NarUw4AALCtpg35n0hyfFXV5Mjx/vHjdAAAYB2Y9uw6L0xyZpLPVtVpSb6WZJ8kT0hyYJIj16Q6AABg2aY9T/77quq+Ga52++Qk+yb5SpKPJ3lia+28tSsRAABYjqnPkz8G+aPXsBYAAGAVTNsnHwAA2E5ssSW/qt6W5AWttQvH21vTWmtPWd3SAACAldhad529k+wy3t4nSVv7cgAAgG21xZDfWjt84vYjZlINAACwzabqk19VL6qqO25h2r5V9aLVLQsAAFipaQ+8fXGS/bcw7Y7jdAAAYB2YNuRXttwnf/8k31ydcgAAgG21tbPrPC3J08a7Lclrq+o7i2bbLcmPJXn32pQHAAAs19bOrvODJF8fb1eSbyf5xqJ5rk5yZpLXrH5pAADASmzt7DqnJjk1SarqL5O8rLV20awKAwAAVmZrLfnXaa09fa0LAQAAVsfW+uS/chnraa21569CPQAAwDbaWkv+k5exnpZEyAcAgHVga33yD5plIQAAwOqY9jz5AADAdmKqA2+TpKoqyUOS3D3D+fFvoLXmNJoAALAOTBXyq+r2Sd6b5JAM/e9rnDR5FVwhHwAA1oFpu+u8KsPFsA7IEPAPTXJgkt9L8vkMrfsAAMA6MG13nYcneU6Sr4z3q7X25SR/UFU7ZWjFf8wa1AcAACzTtC35t0lyeWttc5LvJNlnYtpHkjx4tQsDAABWZtqQ/4Uk+463z03yCxPTHpfkG6tZFAAAsHLThvx/TvLo8fbvJ3lSVV1SVV9I8utJ/mzaf1hVR1TV+VV1QVWduMT0H62qj1bVVVX1vEXTvlhVn6qq/6iqTdP+TwAA2JFM1Se/tfaCidtnVtVDkjw+yc2TnNVaO3Oa9VTVhiQnJXlUkkuSnF1Vp7fWzpuY7RsZfjg8fgurOby1dsU0/w8AAHZEU58nf1Jr7ewkZ69g0QckuaC1dlGSVNVbkxyV5LqQ31q7LMllVfVTK6kNAAB2dFNf8baqblZVx1fVG6rq3VV18Dj+KVV1jylXs1+SiyfuXzKOm1ZL8u6qOqeqjl3GcgAAsMOY9mJYd09yVpI9kpyT5BFJbjVOfmiSn0pyzDSrWmJcW2LcljyktXZpVe2T5Kyq+mxr7YNL1HtskmOT5E53utMyVg8AANu/aVvyX53kyxkugPWY3DCsfyDJYVOu55IMF9RasH+SS6dcNq21S8e/lyU5LUP3n6XmO6W1trG1tnHvvfeedvUAANCFaUP+Q5O8vLX2rdy45f1ruf70mjfl7CQHV9VBVbVrkqOTnD7NglW1e1XdauF2hrP9fHrK/wsAADuMaQ+8vTLDmXSWsl+Sb02zktbaNVV1QpJ3JdmQ5I2ttXOr6rhx+slVdYckm5LcOsnmqvqNJIckuV2S06pqoe63tNbeOWX9AACww5g25J+V5IVV9Z4k3xvHtaq6WZJnJzlj2n/YWjtj8fyttZMnbn81Qzeexb6T5D7T/h8AANhRTRvyfzvJh5NckCHwtyQvSnLPJLsmeeKaVAcAACzbVH3yW2sXZ2hFPznDwbcXZuiHf2qS+4+t7wAAwDpwky35VbVbkj9L8obW2u8l+b01rwoAAFixm2zJb61dmeEsOLutfTkAAMC2mvYUmu9LcvhaFgIAAKyOaQ+8PSnJ68fz05+R4dz4NzhffmvtvFWuDQAAWIFpQ/7C+eifOw6TAb/G+xtWsS4AAGCFpg35uuoAAMB2Ytqz6zw1w9l1Prb2JQEAANvC2XUAAKAzzq4DAACdcXYdAADojLPrAABAZ5xdBwAAOjNVyG+tfWCtCwEAAFbHtC35SZKqOjTJYUlum+QbST7UWvv4WhQGAACszFQhfzzg9tQkRyS5JsnXk+yVZENVvTPJk1trP1izKgEAgKlNewrNVyZ5UJKnJNmttbZvhvPmHz2O/8O1KQ8AAFiuaUP+k5I8v7V2amttc5K01ja31k5NcmKSJ69VgQAAwPJMG/L3SHLxFqZdnOTWq1MOAACwraYN+Z9IcnxV1eTI8f7x43QAAGAdmPbsOi9McmaSz1bVaRmueLtPkickOTDJkWtSHQAAsGzTnif/fVV13yQvytD/ft8kX0ny8SRPbK2dt3YlAgAAyzH1efLHIH/0GtYCAACsgi32ya/B46rqnluZ517jPLWleQAAgNna2oG3P5vkr5N8eyvzfGuc55jVLAoAAFi5rYX8E5Kc3Fq7ZEszjNNek+Tpq10YAACwMlsL+fdN8r4p1vH+cV4AAGAd2FrI3znJVVOs46oku6xOOQAAwLbaWsj/QpL7TbGO+yf54qpUAwAAbLOthfy3J/mtqrrDlmYYpz03yamrXRgAALAyWwv5f5TkO0nOqapfq6q7VtUuVbVzVd2lqo5PcnaGs+/88SyKBQAAbtoWL4bVWvtuVT08yclJ/mwLs52W5PjW2nfXojgAAGD5tnrF29ba5UmeVFV3SvKwJPuNk/4zyQdba19e4/oAAIBl2mrIXzCG+b9Z41oAAIBVsLU++QAAwHZIyAcAgM4I+QAA0BkhHwAAOrPskF+DO1bVVAftAgAAszV1yK+qx1bVx5NcmeTLSe49jj+lqp66RvUBAADLNFXIr6pjkpye5LNJjl203OeTPGP1SwMAAFZi2pb8/5Hkj1prT8uNz5d/bpJDVrUqAABgxaYN+XdOctYWpl2Z5NarUw4AALCtpg35Fye57xambUxyweqUAwAAbKtpQ/4bkrx4PMD25uO4qqqfTPI7SV63FsUBAADLN+1pMP8wyQFJ3pTk2nHcR5JsSPIXrbVXr0FtAADACkwV8ltrLcmzqup/JXlkkr2SfCPJ+1prn1vD+gAAgGVa1gWtWmsXJrlwjWoBAABWwVQhv6oetpXJm5N8J8n5rbWrVqUqAABgxaZtyX9/kjZxvxbdT5Irq+r1SZ7bWrs2AADAXEwb8h+Z4Qw7Z2S48u3lSfZOclSSxyb5rST3SPLCJN/LcPEsAABgDqYN+SckeVNr7SWLxr+rql6S5Jdaa4+rqp2T/FKEfAAAmJtpz5P/6CQf2sK0Dyc5fLz9wST7bmtRAADAyk0b8r+R5Ge2MO1nxulJcosk397WogAAgJWbtrvOK5O8uqoOTPKPuXGf/GeP8x2e5OzVLREAAFiOaS+G9edV9Z9JXpDkpHG5a5L8R5InttbeMc768iRXr0WhAADAdKa+GFZr7bQkp1XVThla8S9vrW1eNM8Vq1wfAACwTMu64m2SjMH+a2tQCwAAsAqmDvljf/ynJrl7kt0WT2+t/dyqVQUAAKzYVCG/qu6f5ANJLs4Q8j+ZZI8kBya5JMkFa1QfAACwTNOeQvOPkrw9yb2SVJJntNbukuSwJC3D2XcAAIB1YNqQ/+NJ3pJk4UDb3ZKktfaRJC9N8orVLw0AAFiJaUN+S3J1a60luSzJnSemXZzk4NUuDAAAWJlpQ/55Se463v5okt+sqoOr6s5JfifJhWtRHAAAsHzTnl3nlAwH2SbJC5O8O8lnx/vfT/Kzq1sWAACwUtNe8favJ25/pqrukeRBSW6e5GOttcvWqD4AAGCZbrK7TlXtVlXvrqpHLIxrrX2vtXZWa+10AR8AANaXmwz5rbUrk/xEkg1rXw4AALCtpj3w9vQkj1/LQgAAgNUx7YG370ryR1W1b5Izknwtw2k1r9NaO2OVawMAAFZg2pD/N+PfJ47DYi268wAAwLowbcg/aE2rAAAAVs20p9D80loXAgAArI5pD7xNVd2sqo6vqjeMp9Q8eBz/lPG8+QAAwDowVUt+Vd09yVlJ9khyTpJHJLnVOPmhSX4qyTFrUB8AALBM07bkvzrJl5McmOQxSWpi2geSHLa6ZQEAACs17YG3D03y5Nbat6pq8Vl0vpZk39UtCwAAWKlpW/KvTHLzLUzbL8m3VqccAABgW00b8s9K8sKq2mNiXKuqmyV5doYLZAEAAOvAtN11fjvJh5NckCHwtyQvSnLPJLtm6QtkAQAAczBVS35r7eIk90lycoaDby/M0A//1CT3b619da0KBAAAlmfalvy01r6Z5PfGAQAAWKemasmvqjdV1ZFVNfWPAgAAYD6mPfD2nkn+OclXq+p1VfXfqqpuaiEAAGD2pu2TvzHJXZO8KsnGJO9J8pWq+vOqeuga1gcAACzTtC35aa19obX28tbafZPcI8lrkjwiyfur6strVB8AALBMU4f8Sa2185O8MclfJvlqhgtiAQAA68CyDqStqjskeXKSpyR5YJJvJnlHkreufmkAAMBKTBXyq+q4DMH+sCTfT/IPSf4gybtba9esXXkAAMByTduS/8dJ/ilDK/6ZrbWrJidW1S6ttR+udnEAAMDyTRvy92mt/WByxHgKzcOT/HySJybZa5VrAwAAVmCqkD8Z8Kvq0AzB/ueS3D7JN6JPPgAArBvT9sm/V4Zgf3SSA5NcnWTXJM9NcpJ++QAAsH5s8RSaVXWXqnphVX0qySeSPC/JZ5Ick+TgJJXk3wV8AABYX7bWkn9Bkpbk40l+NcnbW2vfTJKq2mMGtQHAuvLSqnmXsCIvbm3eJQAztrWLYX0pQ2v9vTJc2fbBVbWs8+oDAACzt8WQ31o7KMlDkrwpyU8m+cckX6uq1433NQsAAMA6tLWW/LTWPtpae3aS/ZI8JsNFsJ6U5O/GWX6lqjaubYkAAMBybDXkL2itbW6tndVa++Ukd8hwXvxTkzwhycer6jNrWCMAALAMU4X8Sa21q1tr72itHZ3hPPnHZDhIFwAAWAeWHfIntda+31r729ba41arIAAAYNtsU8gHAADWH6fEXCPb67mUE+dTBgDY3mnJBwCAzgj5AADQGSEfAAA6I+QDAEBnhHwAAOiMkA8AAJ2ZecivqiOq6vyquqCqTlxi+o9W1Uer6qqqet5ylgUAAGYc8qtqQ5KTkhyZ5JAkP19Vhyya7RtJfj3JH69gWQAA2OHNuiX/AUkuaK1d1Fq7Oslbkxw1OUNr7bLW2tlJfrjcZQEAgNmH/P2SXDxx/5Jx3KouW1XHVtWmqtp0+eWXr6hQAADYXs065NcS49pqL9taO6W1trG1tnHvvfeeujgAAOjBrEP+JUkOmLi/f5JLZ7AsAADsMGYd8s9OcnBVHVRVuyY5OsnpM1gWAAB2GDvP8p+11q6pqhOSvCvJhiRvbK2dW1XHjdNPrqo7JNmU5NZJNlfVbyQ5pLX2naWWnWX9AACwPZhpyE+S1toZSc5YNO7kidtfzdAVZ6plAQCAG3LFWwAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADqz87wLAGD5Xlo17xJW7MWtzbsEgO5pyQcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRGyAcAgM4I+QAA0BkhHwAAOiPkAwBAZ4R8AADojJAPAACdEfIBAKAzQj4AAHRm53kXAKvppVXzLmFFXtzavEsAADqiJR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgMzMP+VV1RFWdX1UXVNWJS0yvqnr1OP2TVXW/iWlfrKpPVdV/VNWm2VYOAADbh5meQrOqNiQ5KcmjklyS5OyqOr21dt7EbEcmOXgcDk3y2vHvgsNba1fMqGQAANjuzLol/wFJLmitXdRauzrJW5MctWieo5K8uQ0+luQ2VbXvjOsEAIDt1qxD/n5JLp64f8k4btp5WpJ3V9U5VXXslv5JVR1bVZuqatPll1++CmUDAMD2Y9Yhf6nLkS6+1OfW5nlIa+1+Gbr0PKuqHrbUP2mtndJa29ha27j33nuvvFoAANgOzTrkX5LkgIn7+ye5dNp5WmsLfy9LclqG7j8AAMCEWYf8s5McXFUHVdWuSY5OcvqieU5Pcsx4lp0HJvl2a+0rVbV7Vd0qSapq9ySPTvLpWRYPAADbg5meXae1dk1VnZDkXUk2JHlja+3cqjpunH5ykjOSPDbJBUl+kOTp4+K3T3JaVS3U/ZbW2jtnWT8AAGwPZhryk6S1dkaGID857uSJ2y3Js5ZY7qIk91nzAgEAYDvnircAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgMzvPuwBg+/bSqnmXsGIvbm3eJQDAmtCSDwAAnRHyAQCgM0I+AAB0RsgHAIDOCPkAANAZIR8AADoj5AMAQGeEfAAA6IyQDwAAnRHyAQCgM0I+AAB0RjHpegwAACAASURBVMgHAIDOCPkAANAZIR8AADoj5AMAQGdmHvKr6oiqOr+qLqiqE5eYXlX16nH6J6vqftMuCwAAJDvP8p9V1YYkJyV5VJJLkpxdVae31s6bmO3IJAePw6FJXpvk0CmXBQA68tKqeZewYi9ubd4lsAObdUv+A5Jc0Fq7qLV2dZK3Jjlq0TxHJXlzG3wsyW2qat8plwUAgB3eTFvyk+yX5OKJ+5dkaK2/qXn2m3LZJElVHZvk2PHu96rq/G2oeT26XZIr1mrlL9mOW03W2Jptd9t8q2z32fMZMx9e6/Nhu89er58xd57XP17KrEP+Ult98b6sLc0zzbLDyNZOSXLK8krbflTVptbaxnnXsaOx3efDdp8923w+bPf5sN1nzzafjVmH/EuSHDBxf/8kl045z65TLAsAADu8WffJPzvJwVV1UFXtmuToJKcvmuf0JMeMZ9l5YJJvt9a+MuWyAACww5tpS35r7ZqqOiHJu5JsSPLG1tq5VXXcOP3kJGckeWySC5L8IMnTt7bsLOtfR7rtirTO2e7zYbvPnm0+H7b7fNjus2ebz0A1p3cCAICuuOItAAB0RsgHAIDOCPkAANAZIR8AADoj5AMAwDpSVRu2dR1CPgDr0ni9lLldn75ntiusb621a5Okqm6+0nUI+SyLL4bZsr3Xv9VobeHGqmrnNqqqW867nl5U1Yaqqub82duNqtpp8XeB74b+VdUtq+qvkzxjvL/s53ymF8Ni+1VVG1pr1y7+YvBlsfoW3sgTAefwJHu31t4259KYMPGeuHZ8znZPclVr7YfjdO+NFZjYrtdU1W5JnpPk8Kp6d2vtf827vu1ZVe000Tr440kekeTbSc5P8qnW2ne9bteP8XNl8jm7W5KDWmtneY52CP8tyS8k+W4yZILlrkDI5yYt+mJ4QJJ7J/lGkv9I8uUk14zzbJ5jmV0YWy+vGW9vSHKLJO9Ncm5Vfbq1dt5cC+Q6E++Jp2doadk9w3vhzUne1Fr7zjzr215NbNfjkvzPJC3D5829J98fLF9rbXNV7ZXk1Ul+Lsm1SXYdJ3+gqp7aWvvPuRXIdSZ+bF1bVfsmeVmG0Hfnqvq11tpfzLdCVsNS2Wniuf94kquS3L6qdm+tfX+569ddh5s0fjHsV1XvSPKvSV6b5O+SfHq8HQF/dYytlztV1e8mOTnJc5P8e5I7ZPhFz4xU1Y9V1e23Mv3gqnpPkjdkaDD5aoag/6dJ3lZVe8+m0u1bVe206P7Nq+pPk7wyyUeT/FaSh7XWfknA3zZVdVCSv09yRJI/SfLkJHdO8kdJfiLJn1bVIeO8uoPM0UKrbVX9WpLPZnjOPpjkhCRvn2NpLNOW3ktjmN+8eL5xD35l+BH++ST7riTgJ0I+U6iq+yY5M8OXwKuS/Ox4+8wkz6iql42tQ6zQwpu7qn4yyUVJfjPDl+99k+w1Do+vqkfMq8YdSVU9McknkhxXVbtsYbb/kWRjkt9Nckxr7cgk90zyh0kemeT5W/uRwHVdcxY3ENw7yTOTvCfJb7bW/qa1dvnsq+vSk5M8NEOof2lr7R9baxcneVOGPYZPTPK0qtpVd5D5q6qfTvL/ZWhce0aSE1prr22tXTHfyliOyfdSVe1cVTsvjK+qParqNVV16MQ8G8Zlvpfk6iSHjD/Ql03IJ8l1B/Zs6QDCZyY5JMlLkrystfYPrbVzMrQEfSJD2DlCy8/KjW/2myd5UYZA/1tJntpae3ySn0nyF0nukeSpVXWL+VW6w/hyhl2lv5Rhu9/A+GPrmCSvb639QWvtc+Pr/24Zuj/slOToJHvOquDt0Xg8w05VdVJV/c44+rAkN0/yJ621i5KkqnatqltV1WOr6kFVdeu5Fb2dGn+s/nKSf2utvaK19r2q2rOqfiZDt6jHJbkwydtaa1fPs9YdSQ1u9N1bVbsmeXaSSvKC1tq7WmvfWzSddWyi8e55YzfOZPi6X+iSe8sMe+iPy7B35rfHGa4du/FcmWRTkg1JbrOSGoT8HVhV3a+q/qmq7tpa2zzRF3bfiXn2S/K0JP+7tfa61tp/VdXeVfWkJM9Pcp8k52X44tDys20el6GV7U2ttb9qrV2WJK21Tyb5/SQfyrAX5XHzK7FPVfWzVXVdd6jW2qYkf5Xkjhla83dftMiDxr9/PS5/UJJfSfK3GfbCvCPJg1prn13j0rc7k40BVXXHJP+S4UvuJ8bw/v0M/fB/rqp+rqqel6Gl+eIk/5Tkw0n+ajw+iNFC6+AS4xe29x2S3D3JF6vqZlX1wCQvTvLmJI9J8huttYPHBhxmoK4/g9S1VbV7VT2gqvYZJ++ZZI8kl7fWPjU+Zz9aVQ+rqpOSvLmqXlFVB8zvEbCUhYa4sfHuVhkai55aVfccn+snVdWFSX6/tfaaDF2xrkryiqp6VVXtN3aT3pDkkgzdQPcY172s3O7A2x3bU5I8NsmXkjxrDPcnJdm7qp7eWrsgyZ0yHPz52bHl4EEZWiiPSXJlkme21t44l+q3M+MuuGuXGL9w4M3CuXA/MY7fZeFMLa21/6yqV2QIOb9YVR9ygNzqGIPm25J8q6re11r7yjjpzAzb+2kZ+jG/Z2KxWyT5YZI7VNWdkzw9yVEZ+s4+srX2vnHduyS5ZqGPpR/C133xbUyyT5KDMoTPZyZ5T2vtO1X13iRvSfJr45AMLcx/leQ7GfasPCnJV6rq3xfeIzuqhdfVROvg4zLs4r+ktXbuxGvuiiSfS3Jwkt/L8Dl+lySnZAj4V06s8x6ttc/M8nHsiCaes+dm2MuyV4aW28e11r5WVZ/K0CX2nAxdN+6Y4Tn7boYW/lsl+bGqemFr7RPzeAxcr6r2TPKXGU6U8dLW2tXjGat+J8lpGY7VujTJT2ZorPjw+D3/7qr62QyNeb+Z5P5V9ezxx93F4+ofmOT9yz7+sbVm2MGGDKfkSpJbJjknQ1h/U4YvhiuSvDDDgR5J8iNJvpXk/2T4Yvhyks0Z+ubXovUeMrl+w3XbpSa3SZIDk9xuifmeM27bly+1HZMckOFAxK8n+a15P64ehiQbxr+/n+FAz1tMvq6TPCHJ5Un+efI5yxA+rx2fj29lCJ8nLLH+RyR5zsLrYN6Pdz0MGY41+UGGA/c/v/B6n9xG42fTkUl+Psmjx3G7jH9/JMllSd4x78cy5+24+PPhUUk+M36GXJlhj8jTk+w+Tr9thh9Km8fhA0l+dIn1PnRc/qfm/Ri39yHJ/ZLc5iamnz1u7zMydNd88MTn0l0yHONz3vh+eUeSx2c4VmtjhkD5Xxl+FMz98e7oQ4bumldlOPPggRPjd8vQUPHDJF9L8qtJ7rXE8rcdv4c2j5+NT8hwnNe3k/zuOM+y8tXcN4phPsPEl+kbxhfeNUlel+ThSW4xMd/dkvzbOH3zGHbuvMT6njB+UP3IvB/beh0yHFD47jGgfCnJy5PcY2L6PcZt/OVc/yNrw8RztdvEl/iHktx73o9pex+W+sBMcoeJ27dJ8mfjNv+lJDuP4w8bP4Q3J/mbJLsusZ4jklw6vn92m/djXQ/bdhx/swzH8WweP1ceMI7fZcr1/uq47Cvn/RjXy5Chle/zGVoH/3D8bLksyTcztBAvNOz88vj58pkk+yxax60y7KH9dIa9iTcKIYapn4/K0AVtc4az4dzoB36GPbdnZGhEeFaS/beyrj2S3H6JaS8b/8dT5/2YDcNnXpLDkzxw0fhXjc/TdzM0Ct188vldYj3PHN+nVyY5PUPIf+eKapr3RjHMZxg/OA7L0Br57fEF+KzJ6RO3X5GhteCsJdazZ5Ljk3whQ9/aJT+oeh4ytDruuXj7Lrr/pAx7SS5M8r4k547b/EOL5vvbcfwfj/d3npi2e4Zw/y/jPCfO+7H3MozvhwPG5+gfF0176BiKPpnkrhPj/3x8Hv508gs4Q0v1s8b5P5WxJXpHGcYvuq22NmVooXz/uP1+Zcr13m3crp/LcFD0Xeb9WOe8nTdk+AF6fIbWv/Mz/mAapz9q/FzelOTQcdzCOfI3Z+iO9vgk+2Y4q85rMrQyfjrJkfN+fNv7kGFP1CeS/MZS74cMe6k2Z2yhnRh/ky21SXbJ0OXjvAx7E28178druNFz9OAkh4+3987QgPr/j8/5K8bxOy9aZnKP/8YMP9oX9rydlUU/zKeqY94bwrD2w+IX0sT4PTO0AP33DLuYPp3kluO0676oM3QveW+GHwSvHl98t83Qen9Khu4jH0vy4Hk/1jls2/tk6Hrwj1uY/rQkD8jQHeqsMTBuGIe3j2/e/zkx/93H9W1O8tSMLcAZzm70+vE5uvcYdM7w4X6Tz8+yushk+BH2X0kePzFulyQvGJ+TF2Xc05UhdL5tHP/ZJC/NcN2ID2TYO7ZpR3tPLPqSukeGg/Ofk+HUjXdatE3/+7jt/jZjV6jFz1eGH0yPzhBAP5ihS+G/Jbn/vB/rvIcMXTYuGz8vzs1wLYHJ6bsm+fVxG7984bMiw3EQC10CNo/LfydDq+Ebkuwx78e2PQ+5fs/rhox7ZCfHj7d3Gj/PNye53zhu2j1ZD85wwPR5Sb6S5OcXr98w8+d8cVi/z/jcvivJARPj75Ch4WdzxkaKjF2ztvAaOijJH+f67js32mN8k7XNe+MY1m7IjftsPjLDr8mfWGLevxhfSC+ZXHbixfagDBfAWvhi+F6G3U5XZWjJvNm8H++ctvFBY5j7QBb1vcz1LTVnjB/G9100/W65vvvN/hPjn54haF41BprTkvzfDLv6fjPDj7Ozxi/mW897G6zXIUP3poUfrQt9XLf0g/dm49/DxufjfZNhJ8OPrPdnaOncODF+zwynlj1vXO6rGVvv5v3457jddx8/T64eX8MLnxkXJvnpXN+3fp8k/5ChkeBJW1jXQzK0UF+RofX+mfN+fDPeltcdz5Old+s/f3zN/TDJwxfPl6GV/oMZzkx05KJl75vhlH0vSPI7meg6aJj6+VmqhX6nRc/BfTPsfT100Xx/Oz5vj5nydfCcDD/mvjJ+P7w3yT3nvQ0MN3iO7jTevk2Gi1l+N8mxi+Y7fvw8PH1Lr6PF7/UMJ33YnEXdgKaqa94bxrD2Q4bdhueOofDa8cXy1iSHTcxz4PjhcUWSg8dxu+SGLXM1rusVSf4gw0WA7jrLx7Iehwxnq1iqT/ZuuX53279ObNPK9T+eFg62/btFyz42w5ldvpqhte7fc8PW5bOT/GeSg+b9+NfjkGFvxzVJ/nm8v/OiL96HZrig292WWHahdf45i8Y/PUOr5xuz6MdVhv7M+2Y4KLS7/veLv3S2Mt9eGfZQ/VeGFqifHt8fr8vQLfBLSX5hnHeh/+r3MoT9A5f6Xxn2Nj4oU7Z09jLkhl31FvYe1aK/d8nQWnhNkp/ewnqOyhAm35zxeJPFocKwTc/TAUl+MclttzD9WePnySm5vtFh5/E79AbdZCeWmfze3XP8++QM/bP/PDtYF8D1PoyfcT/IRL/5JD8+fkd/OBPHz2VoBHnn+Nw/etF69py4XRlzRYYf4Vdm0Q/FqWqb98YxrN0wfpA8f/zC/WiGs+M8Ndf3sf/i5AdTkhPHF96bF63nZrELd5pt/adJfnG8v/Al/Mhxm34tE10Scn3r3K1z/a/0h43jFqbtkiE83m/R/zl2nP+1837c63VIcrsMe0E254b9lA/PcOaDH2b40futDLu+f3Rinv3G5T6TiR8BGU5f95YMrdOPnRi/w+wmnwgclRuG0IXX+xPHL7u/yA0PLtsz14edTRl3YY+v7z/L0Or/a4v+1w6zXbeyvSvDQcrvy7An6a2Tr71xnqdl+KF0xhbWcesM13P4ZoYrM8/9cfUyZGixPWN8XT9q4jk7Jcnzx/v7Z2i1vzJDI9nC5/sTMpwB6RNZ4mQW4zy/m+S9E/eX/CFhmPvr4LYZMtXpGffoj9/fC5nqRZno7ZCh8eMHSc6dWP5xGc6WdKO9mrn+AOtHLbu2eW8cw6q9yG7UDSHJvTKc3ePDufGuwoUzhrx94gt61wzdQjbn+gNG7j++QF8y78e4XoYsvYv2sDH8XZgbd9t587hNT1hqPeOH/Q+TbJqYVov/T4bTrZ2YoXXgooz9vYWhpZ+fDC1fm5OcPd5/dIZuCx/O0Hf+T8b7m8cAdcuJdSx8qL5q0bp/KsNZD/41yR3n/VhnuE1vmeFaAW/LDfuY3jXJXhP33zRut4eO9yf3Wu2a4UfS5kycAnb8nPrC+LzsMH3tM5xe9U5bmf6Ycbt8P8OxOJ/L8GNoc4YuNgtdzG6T/9feeYdZVhT9/9ObYWFJskuOkhaQnEHSSpK8IEgUQXIW1BcBFfAVRAERBFREQZJklZ+AApIFBBEUQXkJwpJZWOLGW78/vnX29D1zZ+PM3Jl76/s89dy53eecOaf73O7q6m9VKcJTrZGC4Mes6/UPkfHEQ2a7D/ujBIVv+bt/Itp5raGdrKKPdkC7sn+kPnJaEWjhV2R0TrRDcwra+fopQcvs6X6dl4q/G51TPYt8Bfcgvn3/rG4JZNR4Dtiscl6hg/2Nkn77DvUGpIFoF/N+n3dmes5pemOGdK0gGkKxkjzTX6J8q2glFILu3153qQ82RWjAnZGy+p5P6gXX+MzOXvJ2Ef8x5z/gOSv1p3lbnVIc758jfaJ+AA8xSkfe5i/83A5cbqSc3owoV1OQgjmy2e3RW6XSroVSuSdS3J+gfmdkGRRlpAZcULnOq8BYModGH8x/iugRa3Tnc/QmQVSk3/lEtBnKJ1BMUt/wY4YgS9YUYHkvqy5Ut0MWrF9SOpUPAL5KafHq4IjWSuK//YKXexyNqX5LIMX+GeTbswAKuTga0Z4mAgdS0ni2Qov/p2jgjO99cyizwOkNmaE+vYX6MNM7Up9XYxhwjtcfTEnDWNuPryGF/nRkgPiz9/EtBCWzp/tyPe+PKzqpL/JO5LrAAEoKzupZeaIMMHAx9QaRpYGzkc/Rcyj0bf/K/xpBqavNkp9X0xs0ZNaEjgmWVqO0whfbhjci68EQn6T3RPzNGqIyNFRSkDXiIaTk3EbmaNiuUvlBL4woHleSxZJGyvwjPgkXSk6h6BeW4e9UrltYnTdCicl2bPC/d0DOc7cCuzS7LfqCUDraboSoUh+h6AS5H0qxsF0CURlqZIsnSsfpqyq/tWWYRoKbVhXEMX0FLfw/QAugnyDOfPGe/8Tb7KROrrEWWig8XClfHO1StUW0KGRZf8J/8x3GYUSzrAH7Z2V5rPtX83N9Pigi5nyt2c/XToL40jWklP8b2KJSnweveArRAJfP6hdEPPsxfp2xyKDzpWY/WzsKMoQ+hYw/+e7u+oji/MvK8cU8UuT+2L5SPz8KnvEu8IUG/28Z6mnT1Ug9h9LAd2yGn6fZDRoye0IZFu3XaNvnm660DASuRjzA0WgVOR4p/XtXrrEA9XG++yGLZURb6Nje/4t2Oca4HFapPwJZMn/VoI2fR3Segsoww85vtBE1ZBb6ZHoRSH6IlPxPgC0rdcV536axNf9eLz+w2c/ZxPYt2mgbtJ08BSmoo+m4m7UBop49QBkibiDlomseZPH8HRlnv12EeurSkd4W38PpGP4uD0JO9zWcq40shbkfz0Vef0B27c8gh/x3ZkcpCOm076pW1qI/NkX0qQv9t3Fs8W5Tv6vYHzjBj/ku9T4r/ZCyvzSKxhMUzOb29fLV+RkFanjJf3c/pwx9Woxte3vdQf4991naFOkNN+HBShq9TzOjE8zwszS7MUNmo/OkUL7mg8vbwE6V+oKu855PJv/byXUuBi5s9vP0ZkHRh273tvwF4mEOy+qLAX8E2mKt4RadbGL/kpdf1JmCU/3hh0y3X/KBdHCjOsQb/7u3/TaV/uqXHfMx2v2aizLM4wbIat320SyQIeEJZM1/Bt/Fop53P4ySH/7TyvlzoR2wqRNhO4u/c7cjq3x18XkrWixVw14W7byHt+PvK/WFo9/h3XXf7SZV5YvKwjYrXwbli/kPsEEn/bYC4l+/QxYQIKTH+7ThIor6HfshrhvtmpWNpKTW/oMshCnyoakBVza47hyUSegO7opnmOFnbXZjh8xAJ1X421l5QSf4N3BLVl4oKItThszcosH5I1Fou49QGuW2D6uWKyyV8uO9Hb/DdLLOIQvnu4g7PzArH0AZ8aVDf4TMVr8dixZXN/k7vXal/qhOlKJCyV8FbbffXq1rJ6FxYpbCUjUCWBnxyScjI8LUaDvZ8StRUg8uQZlXt0KUtbEoZviizX7WZouPNaPRIvIKSofMAWhnaQryVRiUnVO8rwt5W96Fop8VC9rhwGrNfrZWEeqV+1WRBfc2tDjbm3qO9SBEpZroCt38xTUq19nH++4G2pD215sEBVKYv9G873U1ZNgYmpUnSv+Kp3FKHdqxfA0t0OdscL2VqUTZ6ZFnbHYjh0yjczo6ei5KPWe4HwqpVgNuKsr8s5iYC57YPbh1AdF59kUWhXdR6McIzVXf9vmgPBTRD17Myor2/RRy1FkyqxtCuZ1ebN3lvMwIY9d1/bQlZZbaF32QLbJ4HpMdN7e/7zUa8yJ38rofVfu/HYSOjuAbIcW8g6Mm2tW6DVH/RlXqivd8S8QZLxJhfYKU1mvIHBLbQbKxIm/fop2Go+AHHyOfqeLYnZAvyV9xR77K+ZvQidUwpMv7b6jPkVO8T55GUbkmed/ltJvF0W7gWBQmMZ9Hir5dGDmov43nhgjp8T4dgBTuGvDlrHw3nHPvxxTZZo/1skK/GoICmBSRlI70vv8zMnBMc/6gB+lYTW/skOm/BIirdxFy2HkG2CurW80n0NdxDj2Z5QAtFH5MmXlyDKUjYgd+foiBeJM/pkxcMjdlBJZtUYSQI4HrvB0nojjVh+GWHRRS819I8RzRyf8J3uXs9dNyPuE+i6xjw9Gi68veV1O8r4qdrV2QMvUSTmfw43dCu2Gv42FJ20kqY806aEH7gSsxNeBa3I+kOB7Fw38fOZ8v4uVVjuncKIb7cSje95rd/Sy9SRq0RweLvP+9KQqPdxeerwFRmy7K2j+P2LES2rUaS0TL6e4+XBplen/Z54TPUe6aXOf9883KOdshOs4teCZzRNM5kpKPvSINoiqF9GjfbuPzx4Mof0phBPoHZebaNZET7huUuT3y3+5WPsdPQNFxHkKG089O4//26Lzf9IYOmU4HKcvmu4i3eQPavl2icszZNIhoQf2q87PIAfFalHDh+OokFDJ1t+TPSEHcImvDfX0irmXyFOLs/QLFs34F2CG71vf8uH2a/VytKJS7VF/Myor++ipyNL+XjBpCGce92N160Sfwt4E9m/1MTWzLIciy9Z5PcucAXwFO8t/Cw2TO38jw8FNkYNg7Kx+MO9y2s1QUgVFoB+MPKHHSKLIoQkih/7a38/GUIfrWRApmDfG8T0KWxXv82B/QgBYQ0nV9CByOFrMnU0/N2QIZCyYha+5yWd18PidPQpG5TgbuaDRHhzSlX3Ojxrey+eB5nzdWqRx/Ig2CMmT163n/TsiutU2zn3Pq/TX7BkKm0TmyBr+JuN3b0Ul4ObT9NwZZjtf1sk63i2iz9PCdtEFhjWnEv98ZWXVvxfn3PuCviWIcH4+2y/tRxjsuQt7tk11nCdrMetnFfdTpIhQ5Mt2NuOELFcdDXQSSm7xPds7OWxdZbyYgi//BiM8/pNnP26x2RdvS+/s7fxkyCOQT4f3ejj+pnLcJWiQ94GPVCv47mJqozY9rSyUUbd8XSvrL/t69iXZJzqgcuyqi5jxDlrjQ3+NLkWW45uc+lb/TIbPdT9MaZ0YDu2fflwfu9L64i9Kaf1nlvMUo46ZP9Lk5wh83t5/zhXc/lKNgossHyKCaB3Io6FVLICV+MvXZ0/Mxcm6Uy2Y8MqT2mshhTb+BkE46Rhaxm/wFXCcr78wxtEiu8mNKekKj49pywu2kjeegzEg4lZOMLDFFop/9pzUJZNf6pR+/XSf1bcXxns1+qeaAGFat9887aEBZyOr39z65oVL/Q2KHBZxGhqzBl/jkNDyr3wftTk32dh6ft7WfdzRlBK/nkIX5L/jCq10F5QO4D1ngTyOLtgI8iWh+X8jKBqKdk0n+fs5fqVsMOe5t0FPP0OpCx2RGeUz0Yg7tV8wNKKnRq8iC/01gSZ+n3/LfwJb5ddEOzSbA1s1+1nYWOvo2Fg7R86Gd+J/42PaTTB+oOs3u7b/ZPzSoK/p7UZye1Zuk6TfQzkLGm29QtzTaJnyoenwn1xqMttTfwpMxNLpuyNT22swH5nMbtZUPzs+g0IudZhxEFrjvILrHpQQFqiv7aDlEr7kTUZ9yK8oQysRL+1bOK35XS/lv6Dc+URcT9whgo2Y/XxPbdT7gTyj0btEmy1EujpahzMJ5F+KuFgl/ft/gescjK9ZvqYSCbEfxd/NstDA6gHrHzL2Q1bCGlP0hWd2SiMf9dq4Yxjje7f11MNodeQTRqjr4UCF/n4cRZXM09f4VBZf77qws5oFeINQbixYBzkM7wKdT5hgajnIKfULnIWvnpqR67jGd/5l6U/83/QbaUaovAPWrzOKlWtQVlN8xg1kgkRNIzZWiBZv9nL1ZXAF8F1EUVij6IWv/gZQxp0+hpOX0d4VoWxRx4VFk6byOCAvYFf2SW+HHIetJEcFgDFpUFcfs423/ICVlJ99C/Zyf98tmP1dvEpRsZzIemaXSZkugREzjfCLMnflf9fbcy8vy8LC9Znu6m9qsOmZPLwHbwdQ7K38WZTGtIQrPPf731yvn7YAMNb+n4nsVMtt9WDXkDEE7sJ8gx/t/Uy5s18vPAY7xuiMq1+iHHP9fJIvCEtK7BCUr+wQt0h7zuT33sSgoundQv5uZ5wDZElERn6YPhT5t+g20syAHrF+hLaOv5C8O4rf+B22Xdwg5l00wgyuT9K3IuWueZj9fs6XRxOyfBR//BB+Yr+jkuNWRte0VsrjrSOkvwqk9SL2zbVjdZrKPSat/zgAAIABJREFUGky+8yPKx1X+GxmEnA7H+SBcRCCZk5J3/yPqnd+W99/Bm2R0t3aR6rtfqRuBFPbnyGJ5++dB/m5/q3LO8j4JFpzwwjm0pd93lKH3SrRdP9UAkNVXM/72y49BiuQJ3m5PogSGcyNDQc3fz6Wz8z/l/+8FIrxiV/bjgAZlGyLl/BvIf2JeSt+qq8koZ/67qJFlOEf+Et9DjuoHoPm61zhctrvgCjqiVk1AuzRbk9GyKn1ZxL4/qNH74scVNN6duvPeu7Qdmn0D7SJkjp5ou7zY+plMmbDqusoEcpWXn5yV5Vb/T/lgdFBWNrQ7n6MvCB25lp9DoUY/XTluIPC4t/9W1XP9+5XeB+dTJv5ZAG3ZbtzZ/w2ZoX7Kt1JHolCC8yOLyZN0jFhxsvfVN3GePrKQ3ut99DdkQf0OSlbTlhFIpvWslIrqzWjxumyl/mpvy3Uq5bsg+tpDZNb8Vhe0yPwILToXq9R909vjLmQR7pDBFO2avIR2/DbNyhdAFuAa8PPKOb2O19sqgnIRjEI7uYfSMTneAEr/qqOy8nXRwvgxFEt9FFLwxwDnNPu52klmZp5FoUpf8t/vwll5B+ozWvT9A1nql/OyuX2u3yi73mea3QYz1V7NvoF2Eh9A+qNMaq8jh6wNEf++8Nj/Or717QPL+2gV+oVMsUkobNM1KGrDLqHENDzuYGQRex8tpt4FLqA+FOBu3u73VfqpsMSd7ZP8R97OjQaHUO5nrv9y5X4hRF+YhJyf/ooU/BuLvqa0jq6InBlfoJ4KsSxyGh3vfTkebb23c0jMJZAivxsl97RwEBtImWK9SLRUvO9f9fIj/fsgFBP/eVdq5gU+1+zn64H2KxZDc1CG2TvRx++V/B0djxKxFaF1Pyaz9Pr5F3vdZyvlSyMr/kteP6onnqtdpDomA7tn/TQRUaLeBA7w+sHZsZ/2ueKvuEKHdnQOpwyRWIRLvJw+RN1oFUHGoH0pd3UbzsGUmerzHZjqznHxW+/vx49HflxHUlr3/1k9p3qd3ipNv4F2EWAVf3nuQaG1zqPear8WCkX3OrBWVv51H5CmIGrCMShJyj9RAqZTaaOkGjPyw/IB+Rc+mP8ecYvPRpbeGrLyDsuOvyVXbLLyFZD15nLvu/M6G0xCZqkvhyGfk+eR9excRMmpAVeQLWr9sz/i4Y/330DOnRyClK8Ngc2b/Ww93I6NOOHH+Pv/PnBeg/pCmT+6Ur4WUmAnokgStyNa2rNUdq5aSWi8eC/euzWQJfBFRLP5AdoB3BMpGwsAB1Iqf1tRLqh+5OXbZtddyJWI2xCV5ylg5Wa3QSsKWqSuhHah7kC7geciak0NOLOT805qVI+42yegBe9a3X3/IQ37ZjPKxdajyGG9MATlSUAT8H0/bgsv72wxUPzWl8h+s5N9rjmXPhx2vOk30C6CrGfPoogLbwBreHlOLfkysgb9ArcOIKrBKOT5X7zY76KETW05yCCa0u+Agzup/7z/QK8gS8zjk3FBNziVMqPtmigO9QSUj2BxFFHkZp8YRgCrNvu5W0WAPZBSvz0Kvbgfzm32SfSfyFq/anZO7pB+DVoM7FitbyehsT9D7ji+CqUz4Y9xJ1qvXxcZDk717zmdcF20IH4BKbY/xalqrSZ5+zHt6GWHod2ma9GW/qHV8xC3eyLalS3Ck+6CHP7+gGhoOyLr/pv5NUK6vF8H+zx6gfdZ3ULKx5m30M5g4VyevwvDEG3jv7iCGNI7BOXkmORj01veT/tVjinGwWKxdm6D6+S7xEOz8oJt8WUy53c64en3dmn6DbSSdPYSUEZmGU0ZU/pTXpZ7by+KePiFspkPOvOiOMlr4AuEdhXkBFdkgRxeqRtEySsekZUX8W83Rw6Hr+JbfV5+DGX0kDfQLsk4fDs3Oy4s+bPXd0NQ5uYaUuZ/UKkfgKzMk4Ez8RCDld/CVt5Xd1LhlLeLUE95WhM5i+2KLM15eL+1KRP2/KVoL+RU/iEdOcnFWDSUXhr3uZva8zS0a7FWpTwPx1q042Q8kpa/r8Ux86Odw6mWe2S1P8vLJqFFwATgDCJ3Rnf25xJIgf8QGdeKGPZFX80J/C9a6H6VcuclH2d29367mfoMxW1nUOhNgnzsxvv4PxotomsoeMmCfkyhc63g78BTlNSrgZV+HomoeMM7+X9Tkyz2RWn6DbSCVBU/5EC4KpnjYFZ3o7+Qx3Vy7vbAayiOdcOXLsTIJtPTKuUDkPL4ESVfr2rtvMDPPSErG4QsmFeibfRLyLj7ITPdP40mzWKC3RLxXWvAdxvUL+d98C5ZPPusfl5kXX4NWL7Zz9rdbTiN+hGI9jHBlcfCH+Fy6jM3DqPMKfAEZR6N/yJL/+LNftZmtjFlQq8aonXsl9Xn7+8eXv8BFb8EysXRwX6dqyv1+yKr4qnAks1+7lYQPFFVo37wv3dAVMCp0VCo3zlfzvvzSTzJW+X8hKKnXU0lilJIU/t9DrQTdpf30Re8HyciA9Kc2bFDs/n+N5XrDERGv78jqnSHBH60wIKu6TfQSoIszM8iC/CHrqR8D4/D7sesiRTQ5/E4rdRnW50b8cgmA8dOb6JvRaGBgtigbhP/4X4ErJK1Y3/kyPkWHaPfFEriRn7u72gcWi1PXtOBEhEy/b7LvucObfkEWljRftxJP++F+OTX0sCxDVmZ+6x1ZQba8RQfBxrGn0dUnMeRD89ZiBayOqIy1fzc3GdhELJWvYcWBQeiBe2LtPliFjjK2+weH7M/QXG1i93Wgsq0AGUIveOp96kqxqXlfOy/miwBW0iX91murA8HFva5s1+lvHCc/DqldTf38TnE639I6aCeX2OGctSE9Fi/J+Rzdy9yei9+dyujnZsaMgBumZ2zNKL01JBf4z6IjvsdtDh4FS3CW3Keb/oN9CWhpHwUL1Yx+M+NVouT0Jbv95GzxjP+Yt1fGZTO8/Kz/XvV0rwRslJeT5YRsZ0EbX0vnrdzVlco68WE++tK/Zle/p1OzlvbJ/JLGrV/9fiQGe6zXInfGS2i7vQBeT/q406vgjJIvk9mjc/6aDji1I5HW7Kp+j9aVRA15FHkn7NhJ8f8wN/xI6jnk67v7VpDnNJ+1PfLjj62TKZMMtaWWWqzd201b4+vATtR+u1cj+cRyM7ZlNL6u6qX5e27HZGAraf6by6fd59H4WCfQDSofPG1MaJ1PkTmk5LVj0C75v8ly3cS0rsF6Vk1fAfGy5ZFutVEH9t2ptTR1kb+dUVUpII+dycwstnP061t1ewb6AuC6AEnApf6S1GNa7y1K403UG+1XwAp+DXgjKx8OLKgvQ+s5mX5IqAfbRxtAXnL14AnK+XVxdAWyGpWjV6xAVqdj0XUkOoiochke1yzn7XVBFiMMlrRvxEX8j0aZGJGIcqmAJd3cq1RyD+ig+9FKwkNdutQ5JaxiI4zb6VuhNfdkJUtgKz5N3tbP9vZ5OXjVTFJPkgLU55msP3XQLuvN/n3RShzL9xZGVsGU+ZruKoYv71uJUQze5+MZhbSpX1VLPa3ReFHx/p4cxElBfBaygXY3GhXZop/Fg7+ubV+Wz/vBiKJZK8WyoX5Kd5nu1K/qFuHMiztm4hJUYQHHoQW4QegcKgb59et6hetIk2/gd4urhCOQ/SPl31yvA9PrIR4XUXSqiWz84qtwbWQwjm+Un+kn3NNs5+xNwrlLshueXv63wsjK+WfkQWuUFZyi+a3vM0fp4yFvCxyznndjx/RU8/Tl4VOeK/VQRFtpf4EWaBPdaUnoa3UB7yfLsEjFng/FguCzRtcf6gP5oc3uw26u11RxKidKSM+LQj83JWT3am3Fq/hbfZdn7g2RrtaH6AFVcOoLZVrzAPs3uw26A3i79lryPCylJct7wpCzcf/0ZTJ8FZH1t8asgB/He3c3uFlFyPlsiWVhmYL8jH5C7Lg70NJqxrqv4nxwM8oFfqRyNj2PA12xvx6x9Hmi91my8z8Xij9aA7x70MQ5aoGvA38FtF5ijlnhWlcq6Up0U2/gd4qlFaZj5ESv3s2yA+tHPuAH7dKJ9cq6DnfrZQXHLKYbDu22TLFJJqVzYOsLte48vOcK0ZFeNE84cUIyiQ2NZ/An0PbdX+nQXbKkOn2yeb+ObiT+kL5vJYKFxn5otyJfCj2odxGHY2scfdRb5FpWXpO9ZlQ5tSav5eLZeWjUBjLv1TKl/aJ7CY/9wXKMJnVdl+5+j9bsU1noy8Ky+D1SGEvouYU1r9D/J19Daf3eflByOjzNjL83I0UyV2b/UytLkghr5FFPkPUna0pI6v9MevL/oguOAHRe6b6wjX7WUIaW9GZfjz7YvflUuRf9I5/vwVREgciHe7vlA71SzW6VqtL02+gtwqyok1AloFFs/KpyRb8c04UWm0cHn4tqyuOXdVftBuRZ3ih4OyKHELWa/bz9kbxPqgha+5SaKX+DlpQ5ZFxitCkL9Mx9fz2aCV/K3LIidjUM98PA7PJc52sfFvgW9n3A/yYw7PzcuXyS64wXU/p1zIYuNDPO6LZz9rN7Tg1LrN/H+3v7ESUN2A76hc6c6DQjjUU5q9QPJdGCmWxi3UHvrNY+X/bIqvmat35XH1dUESuwmF5VKWuiMzxsX9e5uP5ILSYraHF2LBm3HsrCtNX8C5DO1ZFsrx1kSP/m/6+n9zg3MV8DBsHfKHZzxjSsH83Ac6ZwWOXoQx5XUNUrUOp7M77b/Wb+bzVbtL0G+hNQqmU7+8vzg8r9Z05aJ6VH0/FCokoOxNxB1HCgjCj/TFX9iMukvpcRn3M4qLPCi5yZ87MA6n3e+iTiS2a0AfFO1xEobgd+ZQUdIXr8egslPkLzurkGoujHZUJZBFdEN3kfRTdpOWjkSD6UsH5vtvbrZrvoZpt9SXqk4OdhhZM9zW4/jyIlvYcWgws1exn7q2SjR8neH/s6d8PorQOXoOccgs/hhf8nd0L0XTmavZztILQ0Ul8vuzvvPx3SMnfFYUsfcr75SrcSu/HDcr7Bll4P0HRlNrCittbpToXo0g345ED9XTzAAErIjrOx8jZevnOrt/oHWonafoN9EahtOqs4987sywUE8SKlM6FG1dfJkr+/Un+vS1ftlnsi0O97cZQT1kYgCyjhUV4LVcePyFLaNOgL1rWwaab2j+fXG+iXHQ9ixwQ84RiWyL6wu+BhSvXKazQN3sffaZy7Q2a/aw90JaDEFe45orJiTTIpdHgvMOQkeDHlNbLxVAW1Rryg1gfBQjYyb+/BTwGbNrs5+4LQrkbeCWKnFNDEY4Op8xeOy+K+FQjy0oe0uV9sYm/248iY8Jm1O9wHet98K+snzrshvvvZr9snp6bNg8X29sEGTzWQka882dmHkDW+wnAZ7OyhgbUdp7zm34DvU188nwebaNP1+JOaXEreLWPA/t62eIo/uoYfyGXbPbz9TVBfMpx3rYFv7gzTvjZftyf2/lH3VXtXumD9bxtJyAHzw6DMYqKdAeyyOxFx5CzA13xfB8PTdhO/UT9ztR2nRzzaZR9edmsbCm0OPoQJW8plJb1KGlUNUrH24ko2sjQ7n6mVhHE5y7e7/9DOyUjs/o8U2rDsKYh023jYdnfU2PVZ2X9kcP+RBRVqwi+8ApwUHbc51DCwxoZXTCrn8t/QzUUJ79txpi+ImjBfBvajbwCBcKYbwbP7YcMfBcjS/4mRXmzn6s3Sj8CVYxFA/lcyNo4PfT3z/ORkrky8KuU0n9QRrZL0OR8nJm91PW329owsymIWwyi6mBmE/JjUkrFe3wO2kq/0/xXH5g1eLuTUprX/34PbYffiaJYbOb1g7JzXkI85Y9Qds9dimullBZElIiRwLlmNtbr2qafzOxDRC8AZWmcipTSgimlLyIr/LnZcZjZi2ginICSMC3g5Q+b2RfR7+MsZOk/B9F6DjOzj7r1gVoLD6CoW+8AXzWzU83sadD4YmY1ADP72MwebOJ99kmklA4C3k4p7ZeXZ+PMOshRci80Z34ejRX7IIXuhJTSZn7a39BuoQHDU0qfzv7P+mih8A3kfPvzdhpj+hDGobnkY0RX/J2ZvZtSGjC9E82s5n36CYqqs25R3o3323fR7FVGbxOk4N+JlP3RM3jOUGBp/3tH4Ndou/F24PhmP1MrCGUkov39ezX2/VRHzmbfaysIcoi9AjmLL+Zl/ZBV+Rnk9Llw0faUlrl5UFSjiX7MD5FV9Eo0KN9Hg6Q07SLehsXOVOGovzVyvJ2IrJa7Z8cX7TofpYPyvoTVqqv7ZTG0C/sqZRjT8NuZ9fZM1NPxdqQM+5o7nw+j9Kf6KQ0sushhv4ikUvTNcj7P1pDvyS+RUv+Cl/0SD60Z0rR3oOHvh3JXbDlKavQZzGAoy2yuPxTtGm/f7GftzdL0G+iNgjLW1tCWd8MtpHwQQ466/6gMaoPJYruHzHafLEKZ4KLIQdAZ/y5497Pf3sNQrOGPcPpZVlc4KV7l36vRpAYgR91n/bhPXHk6o6fuvzcLsKG3y5PI6vi6K/jfqhzXv/L9s96mjxLOtN3RL3d6v2zd7Hvpy1KZB5fIxuvNOzn+cBSitAac5mUDsvr+3jfvA1/MygegyGt/RZTYJ5Df0MZd+TwhM93/1bwqu7uOtAueIyWr+6LP6Xcyk/4SKM9KywdqmO3+aPYN9EZBiWgm+qCyS65U0tFCMRI5jTyKrJhFeMxQMru+X35AFr2lMyU/ZIbbM02rDVEq8GIAXikrnxf5PdRwfnKjwRaYH23Bb4Bz8EOmtk2RHGwKsjounNU1HEOQ4+5JhTJUXQSEzHafFMadPZt9L31dKENLP0h9JK35gNPJQr76QuAyb/vTOxlLNvL6P9AxTPJg5O+zWHc8S8gsvwPbomzl4yn9hsa6wj/EjxmOdnCmAHvMit5E7LhNU4KT3wBm9hZKgT0XCp22nZfXzJFSGpJS2ha9oAOBE81snJlN9mODB9j1OBnxME90rnhw8KaDzF+hA/xVrqWUVkopzdngkKeQc9PmwI4FX9LM3kPUEUO+KJjZpJTSsJTSqJTSml421sz+ZWYPmXPwA1Oxh3+OM7MvmdlrKaXBKaVkZpP9U5q+ygea2UTkrHYxcIM5nznQZRgD/AiFhQ3MHrZBEYuuMLNXs/LjUZCKQ4oCM/svouy8isaa4dWLmdkDKDLV1ih0Zo5JZjbJzF7p0icIdEBKKc1IfUrpK8iHayzSodZB/fYv9BvbI6U0yMzeBH6DQiufgGhzM4VC5wo0RgpdtDHcofB6pOC/j8LdPYIi72yOts6/iKwIJ5rZr5p0q22FlNJGwKtm9kKz76U3o6Ik9utsQZRSGo0sbt81s1Ma1C+Lsggm4GCfbIu6y5Fj3PdQZsENkJXmj2hbPRZh00BK6Ycoe+dZZvY/KaX+QC3rt0EoJO/6wJ8tHD67FflvJjB9TKu9UkrF3LmGmT1bjEFudPgv2pXapRhPUkrDkC/Pccjp/BfV8SOltCSagz8GdjOzx7rr2QLTR2b0mVwpnw/tVIKiIj2Y1e2DHKtfAb5iZvemlIYAX0f9/w0UmGFSDzxCe6DZWwm9WZBFoQjLmIep+xBxlW8EFm/2fbajEFt0M9pOpyEH8LUq5cUCf00UreINYJkG5/dDacNrwHnAPFndGiiZUw2YhChuFxI8yRntmyFom7pGJWMtsAriG7/s9Z9v9v2GhExLKB0ih6N8GXeRBUKgpKHt6+/01dRTXzdAu4dPV38P2TGnIUfb6eaXCOmWPj4cuN7/zh2oF8r+3sn7d+esbEXgaO/bms8T+buxGqJ2vUCW9C9k9iXoOtOAmb1pZieicIFfQxFCLkHc8E3NbFcze7mJt9i2sNiimyZSSv1TSkcjitPngCur4esAzOxxxIddEG2jV+trKCHNX9D2+5ZZ3d/QAuBAP3c5MzvCwgozQzCz8WgnBOBMgJTS4h5u8FKUBfJxtPi6tTl3GQg0RkppUErp+pTSt4si/1wd+eNcb1m4YyuprFcgH7Y9qA8l+whwOVIID0gpDW7wb08zs0+b2X+69GEC04X3x/7ArimljU07M59PKT2NdKIC6/vnuymloSmlvYELkJHoPbS7c4SZTfDdS8zs78g3aUng627dD3QBgq4TCLQoUkpHIf7jfSjnwxBkCfuZmb2dUhpg4n8vgviu26LF632V6wz2+n3QJHyqmf13WjSgwIzBOayvAiOQ5X4kogE+CxxpZnc28fYCgU6RUhqKdrZBzrWv+/t8MfAVZI1/vnJOMeasBzwE3I8svmO9fgVk5V0F2MnMHu6hxwlMAyml/qZ8J5sj+ubbyCq/HYpudC6Kdf9hSulANF+cA3wKjWfvoHDi12TXHIayRv/Xvy+DjKn/z8x+23NP19oIS/4MonBgnJ7jSSDQbGTOtveirdFbUazpJ1Cc6otTSvNnlrVXgV+hCfu0lNJAv07y+gl+nfeB3ZBFn1DwZx8mK8s2/vV0YAfgaDNbKRT8QG+GKdlakbStsOQuCmyFdqBeqc6XruD3c+X9KuRzcmBW/yyy6A4HTu4kIECgh2Hu5G9mdyM+/WIoFPDRiFt/tSnZH17/GvKv2Av5HC2SK/iOc4Hrs3nmeTM7NBT8rkUo+TMIKzMextZHoFcjU777Id+RDczsFqSc34+iHFzn0aEK/B5FOdgUbZUPLN71lNJ2fs4ZKBxav1jsdh18q/pulO12QTO7oMm3FAjMKH6BjAN7pZRWRTtSSwI3m9nETubLYuw4HvmkHJJSWjGrvwspgD8zs4+779YD00N1nE8pnYroVB8DcwDX+vhFQb1BQRgKx9uTzezUyjVWSCldAOyM+npQpT700i5E0HUCgRaFb6c/B0wANjOzF1NKywMHoGgGHwBfBu4ypRTfFPgxssadi3YA1kWOcm8gPuYAUwjNQBei2A5v9n0EAjOLlNKGyHjwBFLajgdGAQ+630lx3FR6X0bbOQkZDy40s6N6/u4D00NKaS2U5+SPKaV5kZPs1igSziVmdlh1/Eop7Yh2d4Yi/6Kf+d9bIsfcDYBr0SLg9R59oDZDKPmBQAsiC1l3PVLUNzCzMW6hn5RSOgRxJt8Hfmtmh/h5uwJnAcsiq/0QtFD4spnd35SHCQQCvRoppQeQ4vYhyi8zBUWGegrtEt4PvGJmH/jxU5XClNKHKHnWymb2rybcfqATuFHoGeQ/sZ+Z/Z+XzwPcgeLfr25mT+YhgD285uYoUMlSKPLaRJSl+F3gG+6AHehmDGj2DQQCga6HK/gDgMmIP7kSMCaLfLMq2m5NwFech3+Bmd2YUnoWce8XBx4zs4t6/gkCgUAfwh6U8e/PBD6NFMAdXGrAaymlR4GbgEdSSpPcMffzwIRQ8JuHaQRReAXt6h4ObJ9SOt+EcSmlC5H/xPlop3iKU23M/b3+6LvD66CIOxNQiMzLMipo7GB2M8KSHwi0IDJL/gnA94G9zOwaD894Fkov/xvkILclCrP5InAoou9MnsbAHwgEAnVIKZ0LHAP8j5md5dbeVVEI6tX87+WzUx4DtjGzd3r6XgMlppcEzv0lbkXhLw81s0e9fACaQ3YG9jCz67Jz5jOzd6dxzQEWYbB7BKHkBwItjFRmtL0aTbKroMn1MuAGM3vDeZbnoAg8VwGHFdvqgUAgMCPw2OYfod3BlTxSTl4/F0qgtzay9p7X83cZaISU0nJoB+ZYq+T+cRrOYchi/33gdI+sREppC+AGRPtcFtGuNkQ7wfdUKTlZJJ1QPHsIoeQHAi2MlNLWwB9QRtpXUEK3a8zsaa8vLP5zIm7lg51fLRAIBDpHSumLaIy5ycxGF1bisNz2bqSUTkY5VM43s2Mb1C8MXAMsg0Jm3pbVnQWciOLmv4oWcQOQH9f1PXD7gWkglPxAoIXh1rN/+9fDzezmrC7oOIFAoMtQSe42yszuCgW/d6AR/z1LcrUUcpLdAgVp+GuD80cjRf8q5Dj7mpcvijj7u6BADb/1+vHVawR6HqHkBwItjJTSYmjQXQhY3pSRMCbdQCDQLUgprQb8DXjLzEY0+37aHb7wSln40uHAe2Y2sXLcfihb8V1mtn2D6yyJFPylgZPM7JcN6j8xszf9e8wzvQCRdCAQaGGY2SsoZNlCwEZeFgNvIBDoFnhypD+h6CqROK/J8Gg4tZTSch5S+X7gbymlH6WUNsoOvQXx67dzqz1Ql5xqDDA3mkt2SimtnNeb2Utm9qb3eb+YZ3oHQskPBFofj/nnfE29i0Ag0C7Y1sz2MbNaOFn2PKpZY1NKewAPA+sBY5Fz9FHA/0sp7ZBSmtPMxiF/iteA76SU5qtQOudAjtV/QgmttnKfizrKp/d50EB7CULJDwRaH2OAHwHhBBUIBLodEfu8ucioOZ/3jLVfAO4D9gE2RSFNv4ccZM9EYU4Bbgd+DYwETgIW9OusBFyILPnfBL4F/DwWcL0fwckPBFoc04uDHAgEAoHWQkrpAOBSZNwZhXZXHs7q5waOA05BDrXf8KzoK6MFwPbAo8CzwAoooeJXzexn2TVibunlCCU/EAgEAoFAoA8iC4Ncp3CnlJZGVvkNgEfMbH2PeT+VQuWJri5GSvzuZna/l68AnAFsixgfLwOnmNlvsuuHgt8HEHSdQCAQCAQCgT4EV9inUnMaKNwvAecBBqyQUlqiAY3q3yjr+Qg8G7EvGp4F9gKWAz5nZisUCn7maBsKfh9AKPmBQCAQCAQCfQiFwp5S2j2l9NOU0v+klFbP6mvA3cC1wDzAVl5eWPEH+jGP+ynLZOcBTDaz18zsAT9+QKU+0AcQSn4gEAgEAoFAH0JKabGU0h+QEn8Q8F0UtvRLXp/M7G1Ex/kA+EJKaRWvG2xmk/xSC/jnU/n1q5b6CInZNxFKfiAQCAQCgUAvRWFFr2BvRKc5AVgfGA0MBC5KKX06U9L/CvwcOd9+O6U01MwmpJT6p5S2Ao4G/gP8uZsfI9AENHpxAoFAIBAIBAJNROFUW1jRU0rLqdNeAAAGn0lEQVTLAu8hnv2ewM+A87KQmQuhcMk/AHYGMLOPU0qXAlsDuwILpJTGAJOBjYGFgWPM7I0efbhAjyAs+YFAIBAIBAK9BJlza6G875JS+gdKaHUvcCTi2V/hkXUG+alXA9cBO6aUdsku+Rxwgf89CFgMGAI8BKych8UMtBYihGYgEAgEAoFAL0AlyywppaOBbyNF/R1gRWBJRLHZAHjXzKwIaZlS2hq4DHgb2NDMPvTrLIbi4S8CHGxmf8r+R11ozUDrICz5gUAgEAgEAr0AbpmfM6V0dkppNWBL4C7gS2a2LbAh8C/Ex9/EFfucen0HcBWwCnBIVj4GOAdYAtgnpTQCpkbZmRIKfmsiLPmBQCAQCAQCvQQppWORQv404t9vbGbjssRXOyOr/P+Z2cp+TkI6Xc1DaV4KLA6sZ2Yv+DHzAD8BdgQONbMre/zhAj2KsOQHAoFAIBAI9CCSo5PvFwKPASMR7ebDjHePmd0MXAmslFI6wov7Z4mxngCuBz4F7JCdNw44H/gYODKlNLK7ni/QOxCW/EAgEAgEAoEeQkppQBYxZ2AWs37q95TSrkhRf9nMlvS6BPQzsykppbWAW4BhwEIeRWeqw65H2lnCzB6p/O8hKPrOVsA2ZvZ89z9xoFkIJT8QCAQCgUCgm1E4x/rfA1GM+pHAROAfwHVm9mZ2/I0oFOZRZnZhSql/kenW678NnAqcb2bHVuuz46rOvPOa2Xvd85SB3oRQ8gOBQCAQCARmE7kS3+h7Vr4nimc/DFFn5vOqR4GjzexhP24t4B7gdcStf6dirV8EuBlYG/iMmf2j+54u0BcRnPxAIBAIBAKB2YBby3MrPdn3ftlx2wMXAc8C+wEroKRUlwLrAJellIb5+Y+hhFfLAN/wS5gr+MnMXgUuB/4CfJJz/AMBCEt+IBAIBAKBwCwhp8J4WMr9gWWBpYD7zez0LIZ9f+A3wHbAZoXFPrvWb4HtgTPN7CQvGw48AsyPQmb+3a+TK/uhyAUaIpT8QCAQCAQCgdlASukM4BjgE2AystQnYLSZvePHLAc8DjxjZut4WT+ki01JKa0B3Imy0S5hZm/7MUeiqDg3mtlunfz/hnz8QHsj6DqBQCAQCAQCs4CU0toppQeQgn89cBiwgpltDuxQKPiOT4AJwLsesx4zqxXKuZn9DfgtUvKnKvNmdgHwDLCr8/Q7IBT8QCMMmP4hgUAgEAgEAoEcKaXBwCnAZ4AzgMuK6DhO4/mgQqcZBLwBrAnMDYzLqDxFWM27EVc/+XUGmdlE4CuIovNYTz5joG8jLPmBQCAQCAQCM4jMkfYolGzqdDM7K1PwU5aYaion2mPS34/49UcX9b4gmOyHLeufE72++HzAzB4M59rAzCA4+YFAIBAIBAIzAY+gcycKXznSzF6cFi++qPMss/ciRX934HYz+9CV9zWBq4EPgVFmNrZHHibQsgglPxAIBAKBQGAmkFJaFXgQeMrMNpzBcwpqzjHAtxFl+jrgRmAVtCuwJnACCrNJnsQqEJhZhJIfCAQCgUAgMBNIKS0FPA/8B9jCzMZM5/g82+0Q4POIx7+CHzLFr3Wsmd3RTbcdaDOE420gEAgEAoHAzCEB/wTmQjz6aSr5bsGfG5jHzF4BbkgpPQKsCMwLfGBmt029eBZ/PxCYVYSSHwgEAoFAIDBzeBN4Bdga2Dil9KiZfVI9qBJd5xhgEz8HM3sZeLly/AAzmxwKfqArENF1AoFAIBAIBGYCZvYRcpIFOBBYq4h8kzJkFJ31gSOAySmlQZ1Fycmi7AQCs41Q8gOBQCAQCARmEmZ2OXAbsDTwNWDjrM6cojNnSmlX4GLgHeB/zGyihUNkoAcQdJ1AIBAIBAKBWcO3EKd+e2BkSulY4J8ppTeAUcCmwN4o0+1xZvZk0+400HaI6DqBQCAQCAQCs4iU0urAScBuXmTAB4gt0R+4CTjGzN5pzh0G2hWh5AcCgUAgEAjMBlJKcwBbABuisJivAe8C15nZU35Mf6AWVJ1ATyGU/EAgEAgEAoFZRCWCTsN6pG9FxJxAjyIcbwOBQCAQCARmEbmCn1Lq559FpJ1+7oMbCn6gxxGW/EAgEAgEAoFAoMUQlvxAIBAIBAKBQKDFEEp+IBAIBAKBQCDQYgglPxAIBAKBQCAQaDGEkh8IBAKBQCAQCLQYQskPBAKBQCAQCARaDKHkBwKBQCAQCAQCLYZQ8gOBQCAQCAQCgRZDKPmBQCAQCAQCgUCLIZT8QCAQCAQCgUCgxfD/AZaO54xNxkynAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "dict_corr_avg = {'Concrete': 0.1044, 'Housing': 0.05382, 'Automobile': 0.0080, 'Elevators': 0.0844, 'Abalone': 0.1079, \n",
    "                 'Kinematics': 0.00078, 'Computer Activity': 0.2617}\n",
    "\n",
    "\n",
    "# 'Ailerons': 0.3217, , 'US Dataset': 0.1067\n",
    "\n",
    "# 'Italian AQ': 0.0720, 'Beijing AQ': 0.0131,\n",
    "\n",
    "width = 0.5\n",
    "\n",
    "plt.figure(figsize=(12,12))\n",
    "\n",
    "\n",
    "plt.bar(dict_corr_avg.keys(), dict_corr_avg.values(), width, color='maroon')\n",
    "plt.xticks(rotation=30, fontsize = 20)\n",
    "# plt.yticks(np.arange(0, 120, 20), fontsize = 20)\n",
    "\n",
    "# plt.xlabel(\"Datasets\", fontsize = 12)\n",
    "plt.ylabel(\"Average Correaltion\", fontsize = 15)\n",
    "\n",
    "plt.savefig(\"Img_Avg_Correlation\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.0029805120366832253, 0.030569354222392053, 0.20565533053114254, 0.3225066870462362, 0.3989300726022163, 0.4123041650745128, 1.0]\n"
     ]
    },
    {
     "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": [
    "from matplotlib import pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "clist = [0.00078, 0.0080, 0.05382, 0.0844, 0.1044, 0.1079, 0.2617]\n",
    "clist_norm = [float(i)/max(clist) for i in clist]\n",
    "print(clist_norm)\n",
    "# a = [1,2,5,6,9,11,15,17,18]\n",
    "\n",
    "# plt.hlines(0,0,1)  # Draw a horizontal line\n",
    "# plt.xlim(0,1)\n",
    "# plt.ylim(0,0.5)\n",
    "\n",
    "# y = np.ones(np.shape(clist_norm))   # Make all y values the same\n",
    "# plt.plot(clist_norm, y,'|', ms = 40, linewidth = 40.0)  # Plot a line at each location specified in a\n",
    "# plt.axis('off')\n",
    "# plt.show()\n",
    "\n",
    "\n",
    "plt.figure()\n",
    "# a = [1,2,5,6,9,11,15,17,18]\n",
    "plt.vlines(1,0,1)  # Draw a vertical line\n",
    "clabels = ['Kinematics', 'Automobile', 'Housing', 'Elevators', 'Concrete', 'Abalone', 'C.Activity']\n",
    "# plt.yticks(ticks = clist_norm, )\n",
    "plt.eventplot(clist_norm, orientation='vertical', colors='b', linewidths = 1.5)\n",
    "# plt.annotate('Kinematics', # this is the text\n",
    "#                  (1.0,0.00078), # this is the point to label\n",
    "#                  #textcoords=\"offset points\", # how to position the text\n",
    "#                  #xytext=(0,10), # distance from text to points (x,y)\n",
    "#                  ha='right')\n",
    "plt.axis('off')\n",
    "plt.savefig(\"complexity_line_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Complex: Concrete, Housing, CompAct, Kinematics\n",
    "Simple: Ailerons, Elevators, Auto"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "import networkx as nx\n",
    "\n",
    "G = nx.DiGraph()\n",
    "G.add_edge(\"x\", \"a\", capacity=3.0)\n",
    "G.add_edge(\"x\", \"b\", capacity=1.0)\n",
    "G.add_edge(\"a\", \"c\", capacity=3.0)\n",
    "G.add_edge(\"b\", \"c\", capacity=5.0)\n",
    "G.add_edge(\"b\", \"d\", capacity=4.0)\n",
    "G.add_edge(\"d\", \"e\", capacity=2.0)\n",
    "G.add_edge(\"c\", \"y\", capacity=2.0)\n",
    "G.add_edge(\"e\", \"y\", capacity=3.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "cut_value, partition = nx.minimum_cut(G, \"x\", \"y\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "({'a', 'c', 'x'}, {'b', 'd', 'e', 'y'})"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "partition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "######################Library import ###########################################\n",
    "from two_TrAdaBoostR2 import TwoStageTrAdaBoostR2 ###Import this for STrAdaBoost.R2\n",
    "# from TwoStageTrAdaBoostR2 import TwoStageTrAdaBoostR2 ###Import this for two-stage TrAdaBoost.R2\n",
    "\n",
    "import pandas as pd\n",
    "import sys\n",
    "import numpy as np\n",
    "from pandas import DataFrame\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.ensemble import AdaBoostRegressor\n",
    "\n",
    "from keras.models import Sequential, load_model, Model\n",
    "from keras.layers import Input, Dense, Activation, Conv2D, Dropout, Flatten\n",
    "from keras import optimizers, utils, initializers, regularizers\n",
    "import keras.backend as K\n",
    "\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.preprocessing import StandardScaler #Importing the StandardScaler\n",
    "\n",
    "from itertools import combinations\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from scipy.stats.stats import pearsonr\n",
    "from math import sqrt\n",
    "\n",
    "#Geo plotting libraries\n",
    "import geopandas as gdp\n",
    "from matplotlib.colors import ListedColormap\n",
    "# import geoplot as glpt\n",
    "\n",
    "import xgboost as xgb\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.linear_model import Ridge\n",
    "from sklearn import linear_model\n",
    "from keras.models import Sequential\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "from sklearn.ensemble import GradientBoostingRegressor\n",
    "\n",
    "from sklearn.model_selection import KFold\n",
    "import matplotlib.lines as mlines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "'Concrete': 0.1044, 'Housing': 0.05382, 'Automobile': 0.0080, 'Ailerons': 0.3217, 'Elevators': 0.0844, 'Abalone': 0.1079, \n",
    "                 'Kinematics': 0.00078, 'Computer Activity': 0.2617 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x936 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "################## This is to plot the RMS Error and R2 Errors for the regression datasets###############\n",
    "\n",
    "# Adaboost_RMSE_LM = [0.26038751897237, 5.0376683815286, 8.4791655169811, 0.0033471125695071, 3.0407609461728] #, 3.9425108551414]\n",
    "# AdaboostX_RMSE_LM = [0.26023114662381, 4.378267959749, 7.919944333439, 0.0033427910557569, 3.0432585402816] #, 3.9366965210839]\n",
    "# Tradaboost_RMSE_LM = [0.22266832507436, 5.3923052799087, 8.4363662781185, 0.0043199281964726, 3.1512572784609] #, ]\n",
    "# GBR_RMSE_LM = [0.30639586904018, 6.931528128514, 8.2005217365166, 0.0042351215057175, 3.0554786083319] #, 4.0399359816106]\n",
    "# GBRX_RMSE_LM = [0.25651722138106, 6.6421644634198, 7.7948724964471, 0.0029130548672265, 3.069247399311] #, 3.1676330049544]\n",
    "# STradaboost_RMSE_LM = [0.24019505235318, 4.7747152432431, 7.6403915674999, 0.0029673107729298, 3.033956101467] #, 3.6857811572451]\n",
    "\n",
    "Adaboost_RMSE_LM = [0.84984018, 0.72677601, 1., 0.77480745, 0.96493579]\n",
    "AdaboostX_RMSE_LM = [0.84932981, 0.63164541, 0.93404762, 0.77380709, 0.96572837]\n",
    "Tradaboost_RMSE_LM = [0.7267341,  0.77793889, 0.99495242, 1., 1.]\n",
    "GBR_RMSE_LM = [1., 1., 0.96713783, 0.9803685, 0.9696062]\n",
    "GBRX_RMSE_LM = [0.83720849, 0.95825399, 0.91929713, 0.67432946, 0.97397551]\n",
    "STradaboost_RMSE_LM = [0.78393698, 0.6888402, 0.90107824, 0.68688891, 0.96277639]\n",
    "\n",
    "\n",
    "Adaboost_R2_LM = [0.22271358624651, 0.57100313354671, 0.43788078031171, 0.73176163971293, 0.14354235486685] #, 0.97026804188298]\n",
    "AdaboostX_R2_LM = [0.22297821480852, 0.61166393030941, 0.47400418460823, 0.73276743804607, 0.14185048998762] #, 0.97029209804207]\n",
    "Tradaboost_R2_LM = [0.39414930585469, 0.32140504467159, 0.4076648365164, 0.55183056227661, 0.064264917777793] #, ]\n",
    "GBR_R2_LM = [0.070113345388861, 0.65244131524098, 0.52741113186203, 0.57094634558134, 0.013838183023638] #, 0.96612489398277]\n",
    "GBRX_R2_LM = [0.25061808483802, 0.30896214561578, 0.50210487214982, 0.79740507883259, 0.042394535657282] #, 0.97682152847771]\n",
    "STradaboost_R2_LM = [0.24029499066992, 0.59803600978946, 0.51622360735125, 0.90838470486655, 0.055706541248261] #, 0.96874566723622]\n",
    "\n",
    "# Adaboost_R2_LM = [0.56504878, 0.8751793, 0.83024562, 0.80556359, 1.]\n",
    "# AdaboostX_R2_LM = [0.56572018, 0.9375003, 0.89873754, 0.80667082, 0.98821348]\n",
    "# Tradaboost_R2_LM = [1., 0.49261909, 0.77295455, 0.60748553, 0.447707]\n",
    "# GBR_R2_LM = [0.17788524, 1., 1., 0.62852924, 0.09640488]\n",
    "# GBRX_R2_LM = [0.63584556, 0.47354779, 0.95201796, 0.8778275, 0.29534513]\n",
    "# STradaboost_R2_LM = [0.60965474, 0.91661272, 0.97878785, 1., 0.38808435]\n",
    "\n",
    "\n",
    "######################################################################################################################################################\n",
    "\n",
    "# Adaboost_RMSE_MH = [0.24822159356433, 4.4410804054167, 0.51690158560036, 0.0033471526982521, 6.8975709686383] #, 2.6993846190505 ]\n",
    "# AdaboostX_RMSE_MH = [0.24843890072296, 4.3803926480635, 7.9216075940794, 0.0033440852361765, 6.9008303649283] #, 2.772286253508]\n",
    "# Tradaboost_RMSE_MH = [0.28774892734453, 5.360856013225, 7.8178471482743, 0.004323817446664, 7.9896066558025] #, 3.6570377514222]\n",
    "# GBR_RMSE_MH = [0.19760513310911, 6.5400464761932, 7.8496330309718, 0.0042404945766187, 7.9067484042877] #, 2.0654797687271]\n",
    "# GBRX_RMSE_MH = [0.16083149791428, 6.492341248115, 7.830640174917, 0.0029130548672265, 7.8103578223214] #, 2.1210128353453]\n",
    "# STradaboost_RMSE_MH = [0.22068285334649, 4.32183291862, 3.139327044971, 0.0029673107729298, 7.1043595187502] #,3.4653868980023]\n",
    "\n",
    "Adaboost_RMSE_MH = [0.86263256, 0.67905946, 0.06525211, 0.7741198, 0.86331797]\n",
    "AdaboostX_RMSE_MH = [0.86338776, 0.66978005, 1., 0.77341037, 0.86372592]\n",
    "Tradaboost_RMSE_MH = [1., 0.81969693, 0.98690159, 1., 1.]\n",
    "GBR_RMSE_MH = [0.68672761, 1., 0.99091415, 0.98072933, 0.98962925]\n",
    "GBRX_RMSE_MH = [0.55892997, 0.99270567, 0.98851655, 0.67372291, 0.97756475]\n",
    "STradaboost_RMSE_MH = [0.7669285, 0.66082603, 0.39629924, 0.68627106, 0.88920016]\n",
    "\n",
    "# RMSE_MH = np.array([Adaboost_RMSE_MH, AdaboostX_RMSE_MH, Tradaboost_RMSE_MH, GBR_RMSE_MH, GBRX_RMSE_MH, STradaboost_RMSE_MH]) \n",
    "# R2_LM = np.array([Adaboost_R2_LM, AdaboostX_R2_LM, Tradaboost_R2_LM, GBR_R2_LM, GBRX_R2_LM, STradaboost_R2_LM]) \n",
    "\n",
    "# print(RMSE_MH)\n",
    "# print(\"\\n\")\n",
    "# RMSE_MH = np.abs(RMSE_MH / RMSE_MH.max(axis=0))\n",
    "# print(RMSE_MH)\n",
    "# print(\"\\n\")\n",
    "# print(RMSE_LM.transpose())\n",
    "\n",
    "\n",
    "Adaboost_R2_MH = [0.2025168562006, 0.53419195857315, 0.40779226443498, 0.73191943931014, 0.13622418680961] #, 0.97916491957264]\n",
    "AdaboostX_R2_MH = [0.20202013885999, 0.60832340309585, 0.47817437316266, 0.73276743804607, 0.14020392063643] #, 0.96615127710936]\n",
    "Tradaboost_R2_MH = [0.088924703524272, 0.41637495514861, 0.47828155811508, 0.55183056227661, 0.037350559520466] #, 0.95498143627792]\n",
    "GBR_R2_MH = [0.42799200850992, 0.27045185288114, 0.48970352059286, 0.57105386267691, 0.096512914820321] #, 0.97493916569013]\n",
    "GBRX_R2_MH = [0.64843684209286, 0.34891021292423, 0.49760784524605, 0.79740507883259, 0.07762729100056] #, 0.97442438161888]\n",
    "STradaboost_R2_MH = [0.32360915619155, 0.56191312539345, 0.8449380519317, 0.90838470486655, 0.072739750944493] #, 0.96386044871966]\n",
    "\n",
    "\n",
    "######################################################################################################################################################\n",
    "\n",
    "# Adaboost_RMSE_LH = [0.16781600843473, 4.5686900822772, 3.9513988575627, 0.0033471125695071, 3.1338169744931] #, 2.6979886274737]\n",
    "# AdaboostX_RMSE_LH = [0.16759077896839, 3.4712586371919, 4.310478178893, 0.0033427910557569, 3.1380187981013] #, 2.6988088627822]\n",
    "# Tradaboost_RMSE_LH = [0.21785422801672, 4.2477320040385, 5.0772773637948, 0.0043199281964726, 3.0262011934292] #, 3.3799966123362]\n",
    "# GBR_RMSE_LH = [0.19506236290905, 4.2256670380595, 3.6369527534751, 0.0042351215057175, 2.86882177727] #, CPM ACTIVITY MISSING]\n",
    "# GBRX_RMSE_LH = [0.16087723874659, 3.6255693434073, 4.1314282073083, 0.0029130548672265, 2.7282202974089] #, 2.3711626466495]\n",
    "# STradaboost_RMSE_LH = [0.18562038457534, 3.5048747299387, 4.2466892458194, 0.0029673107729298, 0.18562038457534] #, 2.8878880022154]\n",
    "\n",
    "Adaboost_RMSE_LH = [0.7703133, 1., 0.77825153, 0.77480745, 0.99866099]\n",
    "AdaboostX_RMSE_LH = [0.76927944, 0.75979298, 0.84897434, 0.77380709, 1.]\n",
    "Tradaboost_RMSE_LH = [1., 0.92974834, 1., 1., 0.96436681]\n",
    "GBR_RMSE_LH = [0.8953802, 0.92491873, 0.71631949, 0.9803685, 0.91421434]\n",
    "GBRX_RMSE_LH = [0.73846278, 0.79356868, 0.81370938, 0.67432946, 0.86940853]\n",
    "STradaboost_RMSE_LH = [0.85203939, 0.7671509, 0.83641073, 0.68688891, 0.05915209]\n",
    "\n",
    "\n",
    "# RMSE_LH = np.array([Adaboost_RMSE_LH, AdaboostX_RMSE_LH, Tradaboost_RMSE_LH, GBR_RMSE_LH, GBRX_RMSE_LH, STradaboost_RMSE_LH]) \n",
    "# R2_LM = np.array([Adaboost_R2_LM, AdaboostX_R2_LM, Tradaboost_R2_LM, GBR_R2_LM, GBRX_R2_LM, STradaboost_R2_LM]) \n",
    "\n",
    "# print(RMSE_LH)\n",
    "# print(\"\\n\")\n",
    "# RMSE_LH = np.abs(RMSE_LH / RMSE_LH.max(axis=0))\n",
    "# print(RMSE_LH)\n",
    "# print(\"\\n\")\n",
    "# print(RMSE_LM.transpose())\n",
    "\n",
    "\n",
    "Adaboost_R2_LH = [0.622283848787, 0.46524886681603, 0.70153328584469, 0.73176163971293, 0.1175469169484] #, 0.97916491957264]\n",
    "AdaboostX_R2_LH = [0.62321540348897, 0.59162714947175, 0.67229581334993, 0.73276743804607, 0.11713087775234] #, 0.97914830369038]\n",
    "Tradaboost_R2_LH = [0.40326889105691, 0.49183709477584, 0.57222257093668, 0.55183056227661, 0.072840090013919] #, 0.96885937244308]\n",
    "GBR_R2_LH = [0.4481770492392, 0.51226199477099, 0.75494538096462, 0.57094634558134, 0.030740325582561] #, CPM ACTIVITY MISSING]\n",
    "GBRX_R2_LH = [0.64843684209286, 0.58024645369689, 0.70013799442279, 0.79740507883259, 0.14618440337337] #, 0.98435182139215]\n",
    "STradaboost_R2_LH = [0.52110318618814, 0.57658831335789, 0.67568321896659, 0.90838470486655, 0.087969472625386] #, 0.97884386193814]\n",
    "\n",
    "#######################################################################################################################################################\n",
    "\n",
    "#################################################### CUMULATIVE #############################################################################################\n",
    "############## ['Concrete D', 'Housing', 'Auto', 'Ailerons', 'Elevators', 'Abalone', 'Kinematics', 'CompAct'] (L + M, H) ##############\n",
    "\n",
    "# Adaboost_RMSE = [24.967402079557, 8.4791655169811, 5.0376683815286, 0.00020325265295854, 0.0033471125695071, 3.0407609461728, 0.26038751897237, 3.9425108551414]\n",
    "# AdaboostX_RMSE = [13.964770123091, 7.919944333439, 4.378267959749, 0.00017441106284697, 0.0033427910557569, 3.0432585402816, 0.26023114662381, 3.9366965210839]\n",
    "# Tradaboost_RMSE = [14.566200369473, 8.4363662781185, 5.3923052799087, 0.00021090634603080816, 0.0043199281964726, 3.1512572784609, 0.22266832507436, 4.496007445782896]\n",
    "# GBR_RMSE = [23.681786685621, 8.2005217365166, 6.931528128514, 0.00018387473306345, 0.0042351215057175, 3.0554786083319, 0.30639586904018, 4.0399359816106]\n",
    "# GBRX_RMSE = [11.269768327076, 7.7948724964471, 6.6421644634198, 0.0001611901826149, 0.0029130548672265, 3.069247399311, 0.25651722138106, 3.1676330049544]\n",
    "# STradaboost_RMSE = [13.127525218737, 7.638244897910103, 5.2392486538122665, 0.0001795412822754839, 0.0036899511808519, 3.1082756610754023, 0.23972975088922271, 3.6865219512798]\n",
    "# STradaboostAS_RMSE = [13.176714036197, 7.6403915674999, 4.7747152432431, 0.0001855532678034472, 0.0029673107729298, 3.033956101467, 0.24019505235318, 3.6857811572451]\n",
    "\n",
    "\n",
    "Adaboost_R2 = [0.42967086924117, 0.43788078031171, 0.57100313354671, 0.73176163971293, 0.14354235486685, 0.22271358624651, 0.97026804188298] #0.74849469749676,\n",
    "AdaboostX_R2 = [0.66902638848907, 0.47400418460823, 0.61166393030941, 0.73276743804607, 0.14185048998762, 0.22297821480852, 0.97029209804207] #0.81306444868016,\n",
    "Tradaboost_R2 = [0.65055250676401, 0.4076648365164, 0.32140504467159, 0.55183056227661, 0.064264917777793, 0.39414930585469, 0.9555425657957418] #0.7283795591963251,\n",
    "GBR_R2 = [0.56210100691435, 0.52741113186203, 0.65244131524098, 0.57094634558134, 0.013838183023638, 0.070113345388861, 0.96612489398277] #0.79328008730152,\n",
    "GBRX_R2 = [0.69179304767866, 0.50210487214982, 0.30896214561578, 0.79740507883259, 0.042394535657282, 0.25061808483802, 0.97682152847771] #0.84035559831767,\n",
    "STradaboost_R2 = [0.69443695471351, 0.5172617493422729, 0.540638454649814, 0.77575084511173, 0.038929530316638615, 0.27288855539318535, 0.96884785010965] #0.8142807738642427,\n",
    "STradaboostAS_R2 = [0.69205154599719, 0.51622360735125, 0.59803600978946, 0.90838470486655, 0.055706541248261, 0.24029499066992, 0.96874566723622] #0.7283795591963251,\n",
    "\n",
    "\n",
    "# RMSE_LM = np.array([Adaboost_RMSE, AdaboostX_RMSE,  Tradaboost_RMSE, GBR_RMSE, GBRX_RMSE, STradaboost_RMSE, STradaboostAS_RMSE]) \n",
    "# R2_LM = np.array([Adaboost_R2, AdaboostX_R2, Tradaboost_R2, GBR_R2, GBRX_R2, STradaboost_R2, STradaboostAS_R2]) \n",
    "\n",
    "# print(R2_LM)\n",
    "# print(\"\\n\")\n",
    "# R2_LM = np.abs(R2_LM / R2_LM.max(axis=0))\n",
    "# print(R2_LM)\n",
    "# print(\"\\n\")\n",
    "# print(RMSE_LM.transpose())\n",
    "\n",
    "Adaboost_RMSE = [1., 1., 0.72677601, 0.77480745, 0.96493579, 0.84984018, 0.87689153] #0.8667516,\n",
    "AdaboostX_RMSE = [0.55932011, 0.93404762, 0.63164541, 0.77380709, 0.96572837, 0.84932981, 0.87559831] #0.82695977,\n",
    "Tradaboost_RMSE = [0.58340873, 0.99495242, 0.77793889, 1., 1., 0.7267341, 1.] #1.,\n",
    "GBR_RMSE = [0.94850824, 0.96713783, 1., 0.9803685, 0.9696062, 1., 0.89856079] #0.8718312,\n",
    "GBRX_RMSE = [0.45137929, 0.91929713, 0.95825399, 0.67432946, 0.97397551, 0.83720849, 0.70454354] #0.76427374,\n",
    "STradaboost_RMSE = [0.52578659, 0.90082507, 0.75585766, 0.85416956, 0.98636049, 0.78241835, 0.81995459] #0.8512844,\n",
    "STradaboostAS_RMSE = [0.52775671, 0.90107824, 0.6888402, 0.68688891, 0.96277639, 0.78393698, 0.81978983] #0.87978987,\n",
    "\n",
    "\n",
    "# Adaboost_R2 = [0.61873272, 0.83024562, 0.8751793, 0.80556359, 1., 0.56504878, 0.99329101] #0.89068806,\n",
    "# AdaboostX_R2 = [0.96340839, 0.89873754, 0.9375003, 0.80667082, 0.98821348, 0.56572018, 0.99331564] #0.96752428,\n",
    "# Tradaboost_R2 = [0.93680571, 0.77295455, 0.49261909, 0.60748553, 0.447707, 1., 0.97821612] #0.8667516,\n",
    "# GBR_R2 = [0.80943418, 1., 1., 0.62852924, 0.09640488, 0.17788524, 0.98904955] #0.94398144,\n",
    "# GBRX_R2 = [0.99619273, 0.95201796, 0.47354779, 0.8778275, 0.29534513, 0.63584556, 1.] #1.,\n",
    "# STradaboost_R2 = [1., 0.98075622, 0.82863921, 0.85398933, 0.27120588, 0.69234818, 0.99183712] #0.96897168,\n",
    "# STradaboostAS_R2 = [0.99656497, 0.97878785, 0.91661272, 1., 0.38808435, 0.60965474, 0.99173251] #0.8667516,\n",
    "\n",
    "\n",
    "#######################################################################################################################################################\n",
    "\n",
    "\n",
    "# Adaboost_RMSE = list(map(sum, zip(Adaboost_RMSE_LM, Adaboost_RMSE_LH, Adaboost_RMSE_MH)))\n",
    "# Adaboost_RMSE[:] = [x / 3.0 for x in Adaboost_RMSE]\n",
    "\n",
    "# AdaboostX_RMSE = list(map(sum, zip(AdaboostX_RMSE_LM, AdaboostX_RMSE_LH, AdaboostX_RMSE_MH)))\n",
    "# AdaboostX_RMSE[:] = [x / 3.0 for x in AdaboostX_RMSE]\n",
    "                      \n",
    "# Tradaboost_RMSE = list(map(sum, zip(Tradaboost_RMSE_LM, Tradaboost_RMSE_LH, Tradaboost_RMSE_MH)))\n",
    "# Tradaboost_RMSE[:] = [x / 3.0 for x in Tradaboost_RMSE]                       \n",
    "                       \n",
    "# GBR_RMSE = list(map(sum, zip(GBR_RMSE_LM, GBR_RMSE_LH, GBR_RMSE_MH)))\n",
    "# GBR_RMSE[:] = [x / 3.0 for x in GBR_RMSE]\n",
    "                \n",
    "# GBRX_RMSE = list(map(sum, zip(GBRX_RMSE_LM, GBRX_RMSE_LH, GBRX_RMSE_MH)))\n",
    "# GBRX_RMSE[:] = [x / 3.0 for x in GBRX_RMSE]\n",
    "                 \n",
    "# STradaboost_RMSE = list(map(sum, zip(STradaboost_RMSE_LM, STradaboost_RMSE_LH, STradaboost_RMSE_MH)))\n",
    "# STradaboost_RMSE[:] = [x / 3.0 for x in STradaboost_RMSE]\n",
    "                        \n",
    "                    \n",
    "# Adaboost_R2 = list(map(sum, zip(Adaboost_R2_LM, Adaboost_R2_LH, Adaboost_R2_MH)))\n",
    "# Adaboost_R2[:] = [x / 3.0 for x in Adaboost_R2]\n",
    "                        \n",
    "# AdaboostX_R2 = list(map(sum, zip(AdaboostX_R2_LM, AdaboostX_R2_LH, AdaboostX_R2_MH)))\n",
    "# AdaboostX_R2[:] = [x / 3.0 for x in AdaboostX_R2]\n",
    "                        \n",
    "# Tradaboost_R2 = list(map(sum, zip(Tradaboost_R2_LM, Tradaboost_R2_LH, Tradaboost_R2_MH)))\n",
    "# Tradaboost_R2[:] = [x / 3.0 for x in Tradaboost_R2]\n",
    "                        \n",
    "# GBR_R2 = list(map(sum, zip(GBR_R2_LM, GBR_R2_LH, GBR_R2_MH)))\n",
    "# GBR_R2[:] = [x / 3.0 for x in GBR_R2]\n",
    "                        \n",
    "# GBRX_R2 = list(map(sum, zip(GBRX_R2_LM, GBRX_R2_LH, GBRX_R2_MH)))\n",
    "# GBRX_R2[:] = [x / 3.0 for x in GBRX_R2]\n",
    "                        \n",
    "# STradaboost_R2 = list(map(sum, zip(STradaboost_R2_LM, STradaboost_R2_LH, STradaboost_R2_MH)))\n",
    "# STradaboost_R2[:] = [x / 3.0 for x in STradaboost_R2]                        \n",
    "\n",
    "\n",
    "#xlabels = ['Kinematics', 'Auto', 'Housing', 'Elevators', 'Abalone'] #, 'C.Activity']\n",
    "# xlabels = ['Concrete', 'Housing', 'Auto', 'Ailerons', 'Elevators', 'Abalone', 'Kinematics', 'CompAct']\n",
    "xlabels_RM = ['CN_RM', 'H_RM', 'AU_RM', 'E_RM', 'AB_RM', 'KN_RM', 'CM_RM']\n",
    "xlabels_R2 = ['CN_R2', 'H_R2', 'AU_R2', 'E_R2', 'AB_R2', 'KN_R2', 'CM_R2']\n",
    "# xlabels = ['Kinematics', 'Auto', 'Housing', 'Elevators', 'Concrete', 'Abalone', 'Computer Activity', 'Ailerons']\n",
    "\n",
    "Adaboost = mlines.Line2D([], [], color = 'tomato', marker = 's', linestyle = 'None', markersize = 15, label = 'Adaboost.R2')\n",
    "AdaboostX = mlines.Line2D([], [], color = 'orange', marker = 's', linestyle = 'None', markersize = 15, label = 'Adaboost.R2X')\n",
    "Tradaboost = mlines.Line2D([], [], color = 'darkgreen', marker = 's', linestyle = 'None', markersize = 15, label = 'Two-TrAdaBoost.R2')\n",
    "GBR = mlines.Line2D([], [], color = 'maroon', marker = 's', linestyle = 'None', markersize = 15, label = 'GradBoost')\n",
    "GBRX = mlines.Line2D([], [], color = 'violet', marker = 's', linestyle = 'None', markersize = 15, label = 'GradBoostX')\n",
    "STradaboost_ = mlines.Line2D([], [], color = 'navy', marker = 's', linestyle = 'None', markersize = 15, label = 'STrAdaboost.R2*')\n",
    "STradaboost = mlines.Line2D([], [], color = '#b58500', marker = 's', linestyle = 'None', markersize = 15, label = 'STrAdaboost.R2')\n",
    "RMS = mlines.Line2D([], [], color = 'black', marker = 'o', linestyle = 'None', markersize = 15, label = 'RMS Error')\n",
    "R2 = mlines.Line2D([], [], color = 'black', marker = '*', linestyle = 'None', markersize = 15, label = 'R2 Value')\n",
    "CN = mlines.Line2D([], [], color = 'black', marker = r\"$ {CN} $\", linestyle = 'None', markersize = 15, label = 'Concrete')\n",
    "HO = mlines.Line2D([], [], color = 'black', marker = r\"$ {HO} $\", linestyle = 'None', markersize = 15, label = 'Housing')\n",
    "AU = mlines.Line2D([], [], color = 'black', marker = r\"$ {AU} $\", linestyle = 'None', markersize = 15, label = 'Automobile')\n",
    "EL = mlines.Line2D([], [], color = 'black', marker = r\"$ {EL} $\", linestyle = 'None', markersize = 15, label = 'Elevators')\n",
    "AB = mlines.Line2D([], [], color = 'black', marker = r\"$ {AB} $\", linestyle = 'None', markersize = 15, label = 'Abalone')\n",
    "KN = mlines.Line2D([], [], color = 'black', marker = r\"$ {KN} $\", linestyle = 'None', markersize = 15, label = 'Kinematics')\n",
    "CA = mlines.Line2D([], [], color = 'black', marker = r\"$ {CA} $\", linestyle = 'None', markersize = 15, label = 'Computer Activity')\n",
    "\n",
    "\n",
    "\n",
    "plt.figure(figsize=(13,13))\n",
    "plt.plot(xlabels_RM, Adaboost_RMSE, marker = 'o', color = 'tomato', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "plt.plot(xlabels_R2, Adaboost_R2, marker = '*', color = 'tomato', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "\n",
    "plt.plot(xlabels_RM, AdaboostX_RMSE, marker = 'o', color = 'orange', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "plt.plot(xlabels_R2, AdaboostX_R2, marker = '*', color = 'orange', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "\n",
    "plt.plot(xlabels_RM, Tradaboost_RMSE, marker = 'o', color = 'darkgreen', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "plt.plot(xlabels_R2, Tradaboost_R2, marker = '*', color = 'darkgreen', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "\n",
    "plt.plot(xlabels_RM, GBR_RMSE, marker = 'o', color = 'maroon', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "plt.plot(xlabels_R2, GBR_R2, marker = '*', color = 'maroon', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "\n",
    "plt.plot(xlabels_RM, GBRX_RMSE, marker = 'o', color = 'violet', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "plt.plot(xlabels_R2, GBRX_R2, marker = '*', color = 'violet', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "\n",
    "plt.plot(xlabels_RM, STradaboost_RMSE, marker = 'o', color = 'navy', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "plt.plot(xlabels_R2, STradaboost_R2, marker = '*', color = 'navy', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "\n",
    "plt.plot(xlabels_RM, STradaboostAS_RMSE, marker = 'o', color = '#b58500', markeredgewidth = 4.0, markeredgecolor ='black', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "plt.plot(xlabels_R2, STradaboostAS_R2, marker = '*', color = '#b58500', markeredgewidth = 4.0, markeredgecolor ='black', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "\n",
    "\n",
    "plt.grid(axis = 'x')\n",
    "\n",
    "\n",
    "#labels = ['Kinematics', 'Auto', 'Housing', 'Elevators', 'Abalone'] #, 'C.Activity']\n",
    "# labels = ['Concrete', 'Housing', 'Auto', 'Ailerons', 'Elevators', 'Abalone', 'Kinematics', 'CompAct']\n",
    "# labels = ['CN_RM', 'CN_R2', 'H_RM', 'H_R2', 'AU_RM', 'AU_R2', 'AI_RM', 'AI_R2', 'E_RM', 'E_R2', 'AB_RM', 'AB_R2', 'KN_RM', 'KN_R2', 'CM_RM', 'CM_R2']\n",
    "labels = ['CN_RM', 'HO_RM', 'AU_RM', 'EL_RM', 'AB_RM', 'KN_RM', 'CA_RM', 'CN_R2', 'HO_R2', 'AU_R2', 'EL_R2', 'AB_R2', 'KN_R2', 'CA_R2']\n",
    "# x_val = [0,1,2,3,4,5]\n",
    "# x_val = [0,1,2,3,4,5,6,7]\n",
    "x_val = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]\n",
    "\n",
    "plt.xticks(x_val, labels, rotation = 30, fontsize = 20)\n",
    "# plt.yticks(np.arange(3, 7, 1.0), fontsize = 15)\n",
    "\n",
    "plt.ylabel('Normalized RMS Error/R2 Score', fontsize = 25)\n",
    "# plt.xlabel('Datasets (Decreasing Order of Complexity)', fontsize = 20)\n",
    "\n",
    "plt.legend(handles= [Adaboost, AdaboostX, Tradaboost, GBR, GBRX, STradaboost_, STradaboost, RMS, R2], fontsize = 15, \n",
    "           ncol = 2, handleheight = 2.4, labelspacing = 0.05, loc='lower left') #, CN, HO, AU, EL, AB, KN, CA\n",
    "# plt.legend(handles= [Adaboost, AdaboostX, Tradaboost, GBR, GBRX, STradaboost_, STradaboost], fontsize = 15, loc='lower left')\n",
    "\n",
    "\n",
    "# plt.title('UCI Auto MPG dataset', fontsize = 28)\n",
    "plt.savefig(\"regression_plot_rmse_r2\")\n",
    "plt.show()\n",
    "\n",
    "\n",
    "# plt.figure(figsize=(13,13))\n",
    "# plt.plot(xlabels, Adaboost_R2, marker = 'o', color = 'tomato', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "# plt.plot(xlabels, AdaboostX_R2, marker = 'o', color = 'orange', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "# plt.plot(xlabels, Tradaboost_R2, marker = 'o', color = 'darkgreen', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "# plt.plot(xlabels, GBR_R2, marker = 'o', color = 'maroon', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "# plt.plot(xlabels, GBRX_R2, marker = 'o', color = 'violet', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "# plt.plot(xlabels, STradaboost_R2, marker = 'o', color = 'navy', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "# plt.plot(xlabels, STradaboostAS_R2, marker = 'o', color = '#b58500', linewidth = 4.0, markersize = 20, linestyle = \"None\", alpha = .7)\n",
    "# plt.grid(axis = 'x')\n",
    "\n",
    "# labels = ['Concrete', 'Housing', 'Auto', 'Ailerons', 'Elevators', 'Abalone', 'Kinematics', 'CompAct']\n",
    "# x_val = [0,1,2,3,4,5,6,7]\n",
    "\n",
    "# plt.xticks(x_val, labels, rotation = 30, fontsize = 30)\n",
    "# # plt.yticks(np.arange(3, 7, 1.0), fontsize = 15)\n",
    "\n",
    "# plt.ylabel('R2 Score', fontsize = 30)\n",
    "# # plt.xlabel('Datasets (Decreasing Order of Complexity)', fontsize = 20)\n",
    "\n",
    "# plt.legend(handles= [Adaboost, AdaboostX, Tradaboost, GBR, GBRX, STradaboost_, STradaboost], fontsize = 15)\n",
    "\n",
    "# # plt.title('UCI Auto MPG dataset', fontsize = 28)\n",
    "# # plt.savefig(\"ComplexityPlot_R2_Paper\")\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(153, 19)\n",
      "(2856, 19)\n",
      "(1310, 19)\n"
     ]
    }
   ],
   "source": [
    "########################################## Shorter Individual Datasets (US) ###############################################################\n",
    "\n",
    "US_df_train = pd.read_csv('US_data/Train/Train_3.csv')\n",
    "US_train_droplist = ['cmaq_id', 'cmaq_x', 'cmaq_y', 'Latitude', 'Longitude', 'year', 'month', 'rid', 'clust']\n",
    "US_df_train = US_df_train.drop(US_train_droplist, axis = 1)\n",
    "US_df_train = US_df_train.rename(columns={\"is\": \"ISS\"})\n",
    "print(US_df_train.shape)\n",
    "\n",
    "US_df_transfer = pd.read_csv('US_data/Transfer/Transfer_3.csv')\n",
    "US_transfer_droplist = ['cmaq_id', 'cmaq_x', 'cmaq_y', 'Latitude', 'Longitude', 'year', 'month', 'rid', 'clust']\n",
    "US_df_transfer = US_df_transfer.drop(US_transfer_droplist, axis = 1)\n",
    "US_df_transfer = US_df_transfer.rename(columns={\"is\": \"ISS\"})\n",
    "print(US_df_transfer.shape)\n",
    "\n",
    "US_df_test = pd.read_csv('US_data/Test/Test_3.csv')\n",
    "US_test_droplist = ['cmaq_id', 'cmaq_x', 'cmaq_y', 'Latitude', 'Longitude', 'year', 'month', 'rid']\n",
    "US_df_test = US_df_test.drop(US_test_droplist, axis = 1)\n",
    "US_df_test = US_df_test.rename(columns={\"is\": \"ISS\"})\n",
    "print(US_df_test.shape)\n",
    "\n",
    "target_column = ['pm25_value']\n",
    "US_df_train_y = US_df_train[target_column]\n",
    "US_df_train_X = US_df_train.drop(target_column, axis = 1)\n",
    "\n",
    "US_df_transfer_y = US_df_transfer[target_column]\n",
    "US_df_transfer_X = US_df_transfer.drop(target_column, axis = 1)\n",
    "\n",
    "US_df_test_y = US_df_test[target_column]\n",
    "US_df_test_X = US_df_test.drop(target_column, axis = 1)\n",
    "\n",
    "TF_train_X = pd.concat([US_df_transfer_X, US_df_train_X], sort= False)\n",
    "TF_train_y = pd.concat([US_df_transfer_y, US_df_train_y], sort= False)\n",
    "\n",
    "np_TF_train_X = TF_train_X.to_numpy()\n",
    "np_TF_train_y = TF_train_y.to_numpy()\n",
    "\n",
    "# np_TF_train_X = US_df_train_X.to_numpy()\n",
    "# np_TF_train_y = US_df_train_y.to_numpy()\n",
    "\n",
    "np_US_df_test_X = US_df_test_X.to_numpy()\n",
    "np_US_df_test_y = US_df_test_y.to_numpy()\n",
    "\n",
    "np_TF_train_y_list = np_TF_train_y.ravel()\n",
    "np_US_df_test_y_list = np_US_df_test_y.ravel()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.10676618966867393"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Concrete for the features US: set 1\n",
    "\n",
    "corr_US_df_test_X =  US_df_test_X.corr(method = 'spearman')\n",
    "corr_US_df_test_X\n",
    "\n",
    "US_df_test_X_lsum = np.tril(corr_US_df_test_X).sum() - np.trace(corr_US_df_test_X)\n",
    "US_df_test_X_elems = len(corr_US_df_test_X) * (len(corr_US_df_test_X) - 1) / 2\n",
    "US_df_test_X_lavg = abs(US_df_test_X_lsum / US_df_test_X_elems)\n",
    "US_df_test_X_lavg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.10710321411009029"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Concrete for the features US: set 2\n",
    "\n",
    "corr_US_df_test_X =  US_df_test_X.corr(method = 'spearman')\n",
    "corr_US_df_test_X\n",
    "\n",
    "US_df_test_X_lsum = np.tril(corr_US_df_test_X).sum() - np.trace(corr_US_df_test_X)\n",
    "US_df_test_X_elems = len(corr_US_df_test_X) * (len(corr_US_df_test_X) - 1) / 2\n",
    "US_df_test_X_lavg = abs(US_df_test_X_lsum / US_df_test_X_elems)\n",
    "US_df_test_X_lavg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.10564783161413803"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Get the Spearman Correlation Concrete for the features US: set 3\n",
    "\n",
    "corr_US_df_test_X =  US_df_test_X.corr(method = 'spearman')\n",
    "corr_US_df_test_X\n",
    "\n",
    "US_df_test_X_lsum = np.tril(corr_US_df_test_X).sum() - np.trace(corr_US_df_test_X)\n",
    "US_df_test_X_elems = len(corr_US_df_test_X) * (len(corr_US_df_test_X) - 1) / 2\n",
    "US_df_test_X_lavg = abs(US_df_test_X_lsum / US_df_test_X_elems)\n",
    "US_df_test_X_lavg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'modelRF' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-91-e93743d2a8af>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mimportance_idx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodelRF\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfeature_importances_\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0margsort\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      2\u001b[0m \u001b[0mfig\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msubplots\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfigsize\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;36m15\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m15\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0mimportance_idx\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0;31m# plt.bar(indices, highPower, width=width,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'modelRF' is not defined"
     ]
    }
   ],
   "source": [
    "importance_idx = modelRF.feature_importances_.argsort()\n",
    "fig, axs = plt.subplots(figsize = (15, 15))\n",
    "importance_idx\n",
    "\n",
    "# plt.bar(indices, highPower, width=width, \n",
    "#         color='b', label='Max Power in mW')\n",
    "# plt.bar([i+0.25*width for i in indices], lowPower, \n",
    "#         width=0.5*width, color='r', alpha=0.5, label='Min Power in mW')\n",
    "\n",
    "# plt.xticks(indices+width/2., \n",
    "#            ['T{}'.format(i) for i in range(len(highPower))] )\n",
    "\n",
    "# plt.legend()\n",
    "\n",
    "# plt.show()"
   ]
  },
  {
   "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": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}