TESTING - Regressor.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Import modules"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "#%matplotlib notebook\n",
    "\n",
    "#Basic stuff\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import warnings\n",
    "\n",
    "# ML models\n",
    "from sklearn.model_selection import train_test_split, cross_val_score, cross_val_predict\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.metrics import confusion_matrix, precision_score, recall_score, f1_score\n",
    "from sklearn.metrics import precision_recall_curve, roc_curve, roc_auc_score\n",
    "from sklearn.metrics import make_scorer\n",
    "\n",
    "# Regression metrics\n",
    "from sklearn.metrics import explained_variance_score, median_absolute_error\n",
    "from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score\n",
    "\n",
    "# Initialize tqdm progress bar\n",
    "from tqdm import tqdm, tnrange, tqdm_notebook\n",
    "from tqdm.auto import tqdm\n",
    "tqdm.pandas(tqdm_notebook)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "%reload_ext autoreload\n",
    "%autoreload 2\n",
    "# The averaged random forest model\n",
    "from AverageRF import AverageRF\n",
    "from utils import train_test_split_data, evaluate_model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "def test_model(model, X_train, y_train, X_test, y_test):\n",
    "    \n",
    "    model.fit(X_train,y_train)\n",
    "    y_pred = model.predict(X_test)\n",
    "\n",
    "    # Regression stats\n",
    "    print(f\"Explained Variance Score = {explained_variance_score(y_test,y_pred):.2f}\")\n",
    "    print(f\"Mean Absolute Error      = {mean_absolute_error(y_test, y_pred):.2f}\")\n",
    "    print(f\"Mean SQUARED error       = {mean_squared_error(y_test, y_pred):.2f}\")\n",
    "    print(f\"MEDIAN ansolute error    = {median_absolute_error(y_test, y_pred):.2f}\")\n",
    "    print(f\"R^2 Score                = {r2_score(y_test, y_pred):.2f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "active = pd.read_pickle(\"./active.pkl.bz2\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>PUBCHEM_SID</th>\n",
       "      <th>Curve Class</th>\n",
       "      <th>Log IC50</th>\n",
       "      <th>IC50</th>\n",
       "      <th>R2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>8.290000e+02</td>\n",
       "      <td>829.000000</td>\n",
       "      <td>829.000000</td>\n",
       "      <td>829.000000</td>\n",
       "      <td>829.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>4.170085e+07</td>\n",
       "      <td>1.477443</td>\n",
       "      <td>-4.624427</td>\n",
       "      <td>27.828946</td>\n",
       "      <td>0.974062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>2.642937e+07</td>\n",
       "      <td>0.475725</td>\n",
       "      <td>0.278188</td>\n",
       "      <td>12.781898</td>\n",
       "      <td>0.026871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>8.448520e+05</td>\n",
       "      <td>1.100000</td>\n",
       "      <td>-5.650000</td>\n",
       "      <td>2.238721</td>\n",
       "      <td>0.814713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2.478228e+07</td>\n",
       "      <td>1.100000</td>\n",
       "      <td>-4.800000</td>\n",
       "      <td>15.848932</td>\n",
       "      <td>0.966273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2.672420e+07</td>\n",
       "      <td>1.200000</td>\n",
       "      <td>-4.550000</td>\n",
       "      <td>28.183829</td>\n",
       "      <td>0.982946</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>4.972512e+07</td>\n",
       "      <td>2.100000</td>\n",
       "      <td>-4.400000</td>\n",
       "      <td>39.810717</td>\n",
       "      <td>0.991663</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.031636e+08</td>\n",
       "      <td>2.200000</td>\n",
       "      <td>-4.100000</td>\n",
       "      <td>79.432823</td>\n",
       "      <td>0.999892</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        PUBCHEM_SID  Curve Class    Log IC50        IC50          R2\n",
       "count  8.290000e+02   829.000000  829.000000  829.000000  829.000000\n",
       "mean   4.170085e+07     1.477443   -4.624427   27.828946    0.974062\n",
       "std    2.642937e+07     0.475725    0.278188   12.781898    0.026871\n",
       "min    8.448520e+05     1.100000   -5.650000    2.238721    0.814713\n",
       "25%    2.478228e+07     1.100000   -4.800000   15.848932    0.966273\n",
       "50%    2.672420e+07     1.200000   -4.550000   28.183829    0.982946\n",
       "75%    4.972512e+07     2.100000   -4.400000   39.810717    0.991663\n",
       "max    1.031636e+08     2.200000   -4.100000   79.432823    0.999892"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "active.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Prepare a DataFrame with only the data we need\n",
    "columns = ['PUBCHEM_SID', 'RDKit Fingerprints', 'Morgan Fingerprints','Log IC50']\n",
    "data = active.filter(columns,axis=1)\n",
    "data.reset_index(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_morgan = np.array(list(data['Morgan Fingerprints'])).astype(int)\n",
    "y_data = data['Log IC50'].values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train - Test split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will just separate the indices. This is useful if we want to test with different fingerprints."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train: 746 \t\ty_train: 746\n",
      "X_test:  83 \t\ty_test:  83\n",
      "------------------------------------\n",
      "TOTAL: 829\n"
     ]
    }
   ],
   "source": [
    "# This split will return the indices for X, and the y values\n",
    "X_data = data.index\n",
    "y_data = data['Log IC50'].values\n",
    "X_train_index, X_test_index, y_train, y_test = train_test_split(X_data, y_data, test_size=0.10, shuffle=True)\n",
    "\n",
    "print(\"X_train:\", len(X_train_index), \"\\t\\ty_train:\",len(y_train))\n",
    "print(\"X_test: \", len(X_test_index), \"\\t\\ty_test: \",len(y_test))\n",
    "print(\"-\"*36)\n",
    "print(\"TOTAL:\",len(X_train_index) + len(X_test_index))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Morgan Fingerprints"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((746, 2048), (746,))"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_morgan = np.array(list(data.iloc[X_train_index]['Morgan Fingerprints'])).astype(int)\n",
    "X_test_morgan  = np.array(list(data.iloc[X_test_index]['Morgan Fingerprints'])).astype(int)\n",
    "X_train_morgan.shape, y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(bootstrap=True, criterion='mse', max_depth=None,\n",
       "           max_features='auto', max_leaf_nodes=None,\n",
       "           min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "           min_samples_leaf=1, min_samples_split=2,\n",
       "           min_weight_fraction_leaf=0.0, n_estimators=100, n_jobs=None,\n",
       "           oob_score=False, random_state=None, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "testRF_morgan = RandomForestRegressor(n_estimators=100)\n",
    "testRF_morgan.fit(X_train_morgan,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_morgan = testRF_morgan.predict(X_test_morgan)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = -0.007539566748421933\n",
      "Mean Absolute Error      = 0.16142771084337268\n",
      "Mean SQUARED error       = 0.05440681626506012\n",
      "MEDIAN ansolute error    = 0.11750000000000238\n",
      "R^2 Score                = -0.02711195004315914\n"
     ]
    }
   ],
   "source": [
    "# Regression stats\n",
    "print(\"Explained Variance Score =\", explained_variance_score(y_test,y_morgan))\n",
    "print(\"Mean Absolute Error      =\", mean_absolute_error(y_test, y_morgan))\n",
    "print(\"Mean SQUARED error       =\", mean_squared_error(y_test, y_morgan))\n",
    "print(\"MEDIAN ansolute error    =\", median_absolute_error(y_test, y_morgan))\n",
    "print(\"R^2 Score                =\", r2_score(y_test, y_morgan))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# RDKit Fingerprints"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((746, 2048), (746,))"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_rdkit = np.array(list(data.iloc[X_train_index]['RDKit Fingerprints'])).astype(int)\n",
    "X_test_rdkit  = np.array(list(data.iloc[X_test_index]['RDKit Fingerprints'])).astype(int)\n",
    "X_train_rdkit.shape, y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(bootstrap=True, criterion='mse', max_depth=None,\n",
       "           max_features='auto', max_leaf_nodes=None,\n",
       "           min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "           min_samples_leaf=1, min_samples_split=2,\n",
       "           min_weight_fraction_leaf=0.0, n_estimators=100, n_jobs=None,\n",
       "           oob_score=False, random_state=None, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "testRF_rdkit = RandomForestRegressor(n_estimators=100)\n",
    "testRF_rdkit.fit(X_train_rdkit,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_rdkit = testRF_rdkit.predict(X_test_rdkit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = 0.25\n",
      "Mean Absolute Error      = 0.16\n",
      "Mean SQUARED error       = 0.04\n",
      "MEDIAN ansolute error    = 0.13\n",
      "R^2 Score                = 0.20\n"
     ]
    }
   ],
   "source": [
    "# Regression stats\n",
    "print(f\"Explained Variance Score = {explained_variance_score(y_test,y_rdkit):.2f}\")\n",
    "print(f\"Mean Absolute Error      = {mean_absolute_error(y_test, y_rdkit):.2f}\")\n",
    "print(f\"Mean SQUARED error       = {mean_squared_error(y_test, y_rdkit):.2f}\")\n",
    "print(f\"MEDIAN ansolute error    = {median_absolute_error(y_test, y_rdkit):.2f}\")\n",
    "print(f\"R^2 Score                = {r2_score(y_test, y_rdkit):.2f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SPLIT: 0  [TRAIN: 596 \tTEST: 150] \t Training Time:9.247 seconds.\n",
      "\t  Explained Variance    = 0.3509\n",
      "\t  Mean Absolute Error   = 0.1718\n",
      "\t  Mean Squared Error    = 0.0487\n",
      "\t  Median Absolute Error = 0.1582\n",
      "\t  R^2 Score             = 0.3507 \n",
      "\n",
      "SPLIT: 1  [TRAIN: 597 \tTEST: 149] \t Training Time:9.145 seconds.\n",
      "\t  Explained Variance    = 0.2893\n",
      "\t  Mean Absolute Error   = 0.1987\n",
      "\t  Mean Squared Error    = 0.0620\n",
      "\t  Median Absolute Error = 0.1670\n",
      "\t  R^2 Score             = 0.2810 \n",
      "\n",
      "SPLIT: 2  [TRAIN: 597 \tTEST: 149] \t Training Time:9.304 seconds.\n",
      "\t  Explained Variance    = 0.4468\n",
      "\t  Mean Absolute Error   = 0.1668\n",
      "\t  Mean Squared Error    = 0.0429\n",
      "\t  Median Absolute Error = 0.1520\n",
      "\t  R^2 Score             = 0.4038 \n",
      "\n",
      "SPLIT: 3  [TRAIN: 597 \tTEST: 149] \t Training Time:9.906 seconds.\n",
      "\t  Explained Variance    = 0.3282\n",
      "\t  Mean Absolute Error   = 0.1641\n",
      "\t  Mean Squared Error    = 0.0545\n",
      "\t  Median Absolute Error = 0.1195\n",
      "\t  R^2 Score             = 0.3253 \n",
      "\n",
      "SPLIT: 4  [TRAIN: 597 \tTEST: 149] \t Training Time:9.662 seconds.\n",
      "\t  Explained Variance    = 0.2231\n",
      "\t  Mean Absolute Error   = 0.1988\n",
      "\t  Mean Squared Error    = 0.0661\n",
      "\t  Median Absolute Error = 0.1600\n",
      "\t  R^2 Score             = 0.2065 \n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<AverageRF.AverageRF at 0x29ad2c3d080>"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "avgRF_test = AverageRF(model_type=\"regressor\",n_forests=5,n_estimators=100, verbosity=1)\n",
    "avgRF_test.fit(X_train_rdkit,y_train)"
   ]
  },
  {
   "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>Training Time (s)</th>\n",
       "      <th>Explained Variance</th>\n",
       "      <th>Mean Absolute Error</th>\n",
       "      <th>Mean Squared Error</th>\n",
       "      <th>Median Absolute Error</th>\n",
       "      <th>R^2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9.247281</td>\n",
       "      <td>0.350902</td>\n",
       "      <td>0.171849</td>\n",
       "      <td>0.048680</td>\n",
       "      <td>0.15825</td>\n",
       "      <td>0.350745</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9.144572</td>\n",
       "      <td>0.289271</td>\n",
       "      <td>0.198666</td>\n",
       "      <td>0.062024</td>\n",
       "      <td>0.16700</td>\n",
       "      <td>0.280961</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9.304140</td>\n",
       "      <td>0.446837</td>\n",
       "      <td>0.166818</td>\n",
       "      <td>0.042864</td>\n",
       "      <td>0.15200</td>\n",
       "      <td>0.403803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9.905856</td>\n",
       "      <td>0.328216</td>\n",
       "      <td>0.164125</td>\n",
       "      <td>0.054528</td>\n",
       "      <td>0.11950</td>\n",
       "      <td>0.325295</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9.662188</td>\n",
       "      <td>0.223077</td>\n",
       "      <td>0.198819</td>\n",
       "      <td>0.066142</td>\n",
       "      <td>0.16000</td>\n",
       "      <td>0.206488</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Training Time (s)  Explained Variance  Mean Absolute Error  \\\n",
       "0           9.247281            0.350902             0.171849   \n",
       "1           9.144572            0.289271             0.198666   \n",
       "2           9.304140            0.446837             0.166818   \n",
       "3           9.905856            0.328216             0.164125   \n",
       "4           9.662188            0.223077             0.198819   \n",
       "\n",
       "   Mean Squared Error  Median Absolute Error       R^2  \n",
       "0            0.048680                0.15825  0.350745  \n",
       "1            0.062024                0.16700  0.280961  \n",
       "2            0.042864                0.15200  0.403803  \n",
       "3            0.054528                0.11950  0.325295  \n",
       "4            0.066142                0.16000  0.206488  "
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "avgRF_test_df = pd.DataFrame(avgRF_test.stats)\n",
    "avgRF_test_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_avgRF = avgRF_test.predict(X_test_rdkit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = 0.20\n",
      "Mean Absolute Error      = 0.17\n",
      "Mean SQUARED error       = 0.05\n",
      "MEDIAN ansolute error    = 0.13\n",
      "R^2 Score                = 0.14\n"
     ]
    }
   ],
   "source": [
    "# Regression stats\n",
    "print(f\"Explained Variance Score = {explained_variance_score(y_test,y_pred_avgRF):.2f}\")\n",
    "print(f\"Mean Absolute Error      = {mean_absolute_error(y_test, y_pred_avgRF):.2f}\")\n",
    "print(f\"Mean SQUARED error       = {mean_squared_error(y_test, y_pred_avgRF):.2f}\")\n",
    "print(f\"MEDIAN ansolute error    = {median_absolute_error(y_test, y_pred_avgRF):.2f}\")\n",
    "print(f\"R^2 Score                = {r2_score(y_test, y_pred_avgRF):.2f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = 0.25\n",
      "Mean Absolute Error      = 0.16\n",
      "Mean SQUARED error       = 0.04\n",
      "MEDIAN ansolute error    = 0.13\n",
      "R^2 Score                = 0.20\n"
     ]
    }
   ],
   "source": [
    "# Regression stats\n",
    "print(f\"Explained Variance Score = {explained_variance_score(y_test,y_rdkit):.2f}\")\n",
    "print(f\"Mean Absolute Error      = {mean_absolute_error(y_test, y_rdkit):.2f}\")\n",
    "print(f\"Mean SQUARED error       = {mean_squared_error(y_test, y_rdkit):.2f}\")\n",
    "print(f\"MEDIAN ansolute error    = {median_absolute_error(y_test, y_rdkit):.2f}\")\n",
    "print(f\"R^2 Score                = {r2_score(y_test, y_rdkit):.2f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-4.5545    , -4.73      , -4.445     , -4.9205    , -4.48275   ,\n",
       "       -4.4205    , -4.866     , -4.51725   , -4.546     , -4.6685    ,\n",
       "       -4.49      , -4.703     , -4.893     , -4.5195    , -4.7055    ,\n",
       "       -4.5585    , -4.4045    , -4.6475    , -4.628     , -4.67      ,\n",
       "       -4.6635    , -4.75      , -4.422     , -4.7985    , -4.6385    ,\n",
       "       -4.8825    , -4.562     , -4.789     , -4.5495    , -4.6125    ,\n",
       "       -4.5665    , -4.522     , -4.522     , -4.796     , -4.8605    ,\n",
       "       -4.578     , -4.956     , -4.643     , -4.4885    , -4.62      ,\n",
       "       -4.558     , -4.763     , -4.967     , -4.6485    , -4.5255    ,\n",
       "       -4.6395    , -4.63      , -4.6585    , -4.706     , -4.729     ,\n",
       "       -4.7525    , -4.48266667, -4.777     , -4.63      , -4.541     ,\n",
       "       -4.565     , -4.8495    , -4.4975    , -5.0665    , -4.48575   ,\n",
       "       -4.5465    , -4.455     , -4.4645    , -4.43883333, -4.758     ,\n",
       "       -4.775     , -4.493     , -4.5565    , -4.713     , -4.605     ,\n",
       "       -4.5755    , -4.579     , -4.518     , -4.6255    , -4.462     ,\n",
       "       -4.472     , -4.7705    , -4.62183333, -4.8375    , -4.6835    ,\n",
       "       -4.651     , -4.706     , -4.56725   ])"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_rdkit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = 0.23\n",
      "Mean Absolute Error      = 0.16\n",
      "Mean SQUARED error       = 0.04\n",
      "MEDIAN ansolute error    = 0.12\n",
      "R^2 Score                = 0.22\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import GradientBoostingRegressor\n",
    "gboos_reg = GradientBoostingRegressor(n_estimators=250, max_features='sqrt')\n",
    "test_model(gboos_reg, X_train_rdkit, y_train, X_test_rdkit, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = -846508297461.99\n",
      "Mean Absolute Error      = 157814.48\n",
      "Mean SQUARED error       = 44840127319.62\n",
      "MEDIAN ansolute error    = 106774.01\n",
      "R^2 Score                = -846508466641.59\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import SGDRegressor\n",
    "dgr_reg = SGDRegressor(max_iter=1000, tol=1e-3)\n",
    "test_model(dgr_reg, X_train_rdkit, y_train, X_test_rdkit, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = 0.27\n",
      "Mean Absolute Error      = 0.16\n",
      "Mean SQUARED error       = 0.04\n",
      "MEDIAN ansolute error    = 0.15\n",
      "R^2 Score                = 0.24\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import BayesianRidge\n",
    "bay_reg = BayesianRidge(n_iter=1000, normalize=True)\n",
    "test_model(bay_reg, X_train_rdkit, y_train, X_test_rdkit, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = -1.10\n",
      "Mean Absolute Error      = 0.24\n",
      "Mean SQUARED error       = 0.11\n",
      "MEDIAN ansolute error    = 0.17\n",
      "R^2 Score                = -1.13\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import ARDRegression\n",
    "ard_reg = ARDRegression(n_iter=2000)\n",
    "test_model(ard_reg, X_train_rdkit, y_train, X_test_rdkit, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = 0.33\n",
      "Mean Absolute Error      = 0.15\n",
      "Mean SQUARED error       = 0.04\n",
      "MEDIAN ansolute error    = 0.12\n",
      "R^2 Score                = 0.31\n"
     ]
    }
   ],
   "source": [
    "from sklearn.svm import SVR\n",
    "svr_reg = SVR(kernel=\"rbf\",gamma=\"scale\", epsilon=0.05)\n",
    "test_model(svr_reg, X_train_rdkit, y_train, X_test_rdkit, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Support Vector Regressor')"
      ]
     },
     "execution_count": 215,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "svr_reg.fit(X_train_rdkit, y_train)\n",
    "y_svr = svr_reg.predict(X_test_rdkit)\n",
    "fit_fn = np.poly1d(np.polyfit(y_test,rf_pred,1))\n",
    "fit_lim = [min(y_test),max(y_test)]\n",
    "plt.plot(y_test,y_svr, \"b.\",fit_lim,fit_fn(fit_lim),\"r\")\n",
    "plt.xlabel(\"Exp. pIC50\")\n",
    "plt.ylabel(\"Calc. pIC50\")\n",
    "plt.title(\"Support Vector Regressor\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = 0.24\n",
      "Mean Absolute Error      = 0.16\n",
      "Mean SQUARED error       = 0.04\n",
      "MEDIAN ansolute error    = 0.12\n",
      "R^2 Score                = 0.20\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "rf_reg = RandomForestRegressor(bootstrap=False, criterion='mse',\n",
    "           max_features=\"sqrt\", n_estimators=1000, n_jobs=4, warm_start=False, \n",
    "           random_state=42)\n",
    "test_model(rf_reg, X_train_rdkit, y_train, X_test_rdkit, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [],
   "source": [
    "rf_reg.fit(X_train_rdkit, y_train)\n",
    "rf_pred = rf_reg.predict(X_test_rdkit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Random Forest')"
      ]
     },
     "execution_count": 214,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fit_fn = np.poly1d(np.polyfit(y_test,rf_pred,1))\n",
    "fit_lim = [min(y_test),max(y_test)]\n",
    "plt.plot(y_test,rf_pred, \"b.\",fit_lim,fit_fn(fit_lim),\"r\")\n",
    "plt.xlabel(\"Exp. pIC50\")\n",
    "plt.ylabel(\"Calc. pIC50\")\n",
    "plt.title(\"Random Forest\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = 0.33\n",
      "Mean Absolute Error      = 0.15\n",
      "Mean SQUARED error       = 0.04\n",
      "MEDIAN ansolute error    = 0.12\n",
      "R^2 Score                = 0.29\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import AdaBoostRegressor\n",
    "adab_reg = AdaBoostRegressor(base_estimator=svr_reg, n_estimators=50, loss=\"linear\")\n",
    "test_model(adab_reg, X_train_rdkit, y_train, X_test_rdkit, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = 0.33\n",
      "Mean Absolute Error      = 0.15\n",
      "Mean SQUARED error       = 0.04\n",
      "MEDIAN ansolute error    = 0.12\n",
      "R^2 Score                = 0.29\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import AdaBoostRegressor\n",
    "adab_reg = AdaBoostRegressor(base_estimator=svr_reg, n_estimators=50, loss=\"exponential\")\n",
    "test_model(adab_reg, X_train_rdkit, y_train, X_test_rdkit, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = 0.26\n",
      "Mean Absolute Error      = 0.16\n",
      "Mean SQUARED error       = 0.04\n",
      "MEDIAN ansolute error    = 0.13\n",
      "R^2 Score                = 0.21\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import AdaBoostRegressor\n",
    "adab_reg = AdaBoostRegressor(base_estimator=rf_reg, n_estimators=50, loss=\"exponential\")\n",
    "test_model(adab_reg, X_train_rdkit, y_train, X_test_rdkit, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance Score = 0.11\n",
      "Mean Absolute Error      = 0.17\n",
      "Mean SQUARED error       = 0.05\n",
      "MEDIAN ansolute error    = 0.12\n",
      "R^2 Score                = 0.08\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import AdaBoostRegressor\n",
    "adab_reg = AdaBoostRegressor(base_estimator=bay_reg, n_estimators=50, loss=\"linear\", random_state=42)\n",
    "test_model(adab_reg, X_train_rdkit, y_train, X_test_rdkit, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "UNC-ML (Python 3.7.1)",
   "language": "python",
   "name": "unc"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}