{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DragonNet vs Meta-Learners Benchmark with IHDP + Synthetic Datasets"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {
    "nterop": {
     "id": "36"
    }
   },
   "source": [
    "`Dragonnet` requires `tensorflow`. If you haven't, please install `tensorflow` as follows:\n",
    "```\n",
    "pip install tensorflow\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:02.831394Z",
     "start_time": "2021-06-05T00:41:02.799836Z"
    },
    "nterop": {
     "id": "1"
    }
   },
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:07.549904Z",
     "start_time": "2021-06-05T00:41:02.832846Z"
    },
    "nterop": {
     "id": "2"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.model_selection import train_test_split, StratifiedKFold\n",
    "from sklearn.linear_model import LogisticRegressionCV, LogisticRegression\n",
    "from xgboost import XGBRegressor\n",
    "from lightgbm import LGBMRegressor\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from sklearn.metrics import mean_squared_error as mse\n",
    "from scipy.stats import entropy\n",
    "import warnings\n",
    "\n",
    "from causalml.inference.meta import LRSRegressor\n",
    "from causalml.inference.meta import XGBTRegressor, MLPTRegressor\n",
    "from causalml.inference.meta import BaseXRegressor, BaseRRegressor, BaseSRegressor, BaseTRegressor\n",
    "from causalml.inference.tf import DragonNet\n",
    "from causalml.match import NearestNeighborMatch, MatchOptimizer, create_table_one\n",
    "from causalml.propensity import ElasticNetPropensityModel\n",
    "from causalml.dataset.regression import *\n",
    "from causalml.metrics import *\n",
    "\n",
    "import os, sys\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "warnings.filterwarnings('ignore')\n",
    "plt.style.use('fivethirtyeight')\n",
    "sns.set_palette('Paired')\n",
    "plt.rcParams['figure.figsize'] = (12,8)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {
    "nterop": {
     "id": "3"
    }
   },
   "source": [
    "## IHDP semi-synthetic dataset\n",
    "\n",
    "Hill introduced a semi-synthetic dataset constructed from the Infant Health\n",
    "and Development Program (IHDP). This dataset is based on a randomized experiment\n",
    "investigating the effect of home visits by specialists on future cognitive scores. The data has 747 observations (rows). The IHDP simulation is considered the de-facto standard benchmark for neural network treatment effect\n",
    "estimation methods.\n",
    "\n",
    "The original [paper](https://arxiv.org/pdf/1906.02120.pdf) uses 1000 realizations from the NCPI package, but for illustration purposes, we use 1 dataset (realization) as an example below. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:07.632983Z",
     "start_time": "2021-06-05T00:41:07.551613Z"
    },
    "nterop": {
     "id": "4"
    }
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv(f'data/ihdp_npci_3.csv', header=None)\n",
    "cols =  [\"treatment\", \"y_factual\", \"y_cfactual\", \"mu0\", \"mu1\"] + [f'x{i}' for i in range(1,26)]\n",
    "df.columns = cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:07.781293Z",
     "start_time": "2021-06-05T00:41:07.634456Z"
    },
    "nterop": {
     "id": "5"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(747, 30)"
      ]
     },
     "execution_count": 4,
     "metadata": {
      "nterop": {
       "id": "42"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:08.258383Z",
     "start_time": "2021-06-05T00:41:07.782535Z"
    },
    "nterop": {
     "id": "7"
    },
    "scrolled": true
   },
   "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>treatment</th>\n",
       "      <th>y_factual</th>\n",
       "      <th>y_cfactual</th>\n",
       "      <th>mu0</th>\n",
       "      <th>mu1</th>\n",
       "      <th>x1</th>\n",
       "      <th>x2</th>\n",
       "      <th>x3</th>\n",
       "      <th>x4</th>\n",
       "      <th>x5</th>\n",
       "      <th>...</th>\n",
       "      <th>x16</th>\n",
       "      <th>x17</th>\n",
       "      <th>x18</th>\n",
       "      <th>x19</th>\n",
       "      <th>x20</th>\n",
       "      <th>x21</th>\n",
       "      <th>x22</th>\n",
       "      <th>x23</th>\n",
       "      <th>x24</th>\n",
       "      <th>x25</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>5.931652</td>\n",
       "      <td>3.500591</td>\n",
       "      <td>2.253801</td>\n",
       "      <td>7.136441</td>\n",
       "      <td>-0.528603</td>\n",
       "      <td>-0.343455</td>\n",
       "      <td>1.128554</td>\n",
       "      <td>0.161703</td>\n",
       "      <td>-0.316603</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>2.175966</td>\n",
       "      <td>5.952101</td>\n",
       "      <td>1.257592</td>\n",
       "      <td>6.553022</td>\n",
       "      <td>-1.736945</td>\n",
       "      <td>-1.802002</td>\n",
       "      <td>0.383828</td>\n",
       "      <td>2.244320</td>\n",
       "      <td>-0.629189</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>2.180294</td>\n",
       "      <td>7.175734</td>\n",
       "      <td>2.384100</td>\n",
       "      <td>7.192645</td>\n",
       "      <td>-0.807451</td>\n",
       "      <td>-0.202946</td>\n",
       "      <td>-0.360898</td>\n",
       "      <td>-0.879606</td>\n",
       "      <td>0.808706</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>3.587662</td>\n",
       "      <td>7.787537</td>\n",
       "      <td>4.009365</td>\n",
       "      <td>7.712456</td>\n",
       "      <td>0.390083</td>\n",
       "      <td>0.596582</td>\n",
       "      <td>-1.850350</td>\n",
       "      <td>-0.879606</td>\n",
       "      <td>-0.004017</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>2.372618</td>\n",
       "      <td>5.461871</td>\n",
       "      <td>2.481631</td>\n",
       "      <td>7.232739</td>\n",
       "      <td>-1.045229</td>\n",
       "      <td>-0.602710</td>\n",
       "      <td>0.011465</td>\n",
       "      <td>0.161703</td>\n",
       "      <td>0.683672</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   treatment  y_factual  y_cfactual       mu0       mu1        x1        x2  \\\n",
       "0          1   5.931652    3.500591  2.253801  7.136441 -0.528603 -0.343455   \n",
       "1          0   2.175966    5.952101  1.257592  6.553022 -1.736945 -1.802002   \n",
       "2          0   2.180294    7.175734  2.384100  7.192645 -0.807451 -0.202946   \n",
       "3          0   3.587662    7.787537  4.009365  7.712456  0.390083  0.596582   \n",
       "4          0   2.372618    5.461871  2.481631  7.232739 -1.045229 -0.602710   \n",
       "\n",
       "         x3        x4        x5  ...  x16  x17  x18  x19  x20  x21  x22  x23  \\\n",
       "0  1.128554  0.161703 -0.316603  ...    1    1    1    1    0    0    0    0   \n",
       "1  0.383828  2.244320 -0.629189  ...    1    1    1    1    0    0    0    0   \n",
       "2 -0.360898 -0.879606  0.808706  ...    1    0    1    1    0    0    0    0   \n",
       "3 -1.850350 -0.879606 -0.004017  ...    1    0    1    1    0    0    0    0   \n",
       "4  0.011465  0.161703  0.683672  ...    1    1    1    1    0    0    0    0   \n",
       "\n",
       "   x24  x25  \n",
       "0    0    0  \n",
       "1    0    0  \n",
       "2    0    0  \n",
       "3    0    0  \n",
       "4    0    0  \n",
       "\n",
       "[5 rows x 30 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {
      "nterop": {
       "id": "43"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:08.406084Z",
     "start_time": "2021-06-05T00:41:08.259635Z"
    },
    "nterop": {
     "id": "9"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    0.813922\n",
       "1    0.186078\n",
       "Name: treatment, dtype: float64"
      ]
     },
     "execution_count": 6,
     "metadata": {
      "nterop": {
       "id": "44"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.Series(df['treatment']).value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:08.511617Z",
     "start_time": "2021-06-05T00:41:08.407305Z"
    },
    "nterop": {
     "id": "11"
    }
   },
   "outputs": [],
   "source": [
    "X = df.loc[:,'x1':]\n",
    "treatment = df['treatment']\n",
    "y = df['y_factual']\n",
    "tau = df.apply(lambda d: d['y_factual'] - d['y_cfactual'] if d['treatment']==1 \n",
    "               else d['y_cfactual'] - d['y_factual'], \n",
    "               axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:10.043994Z",
     "start_time": "2021-06-05T00:41:08.605684Z"
    },
    "nterop": {
     "id": "13"
    }
   },
   "outputs": [],
   "source": [
    "p_model = ElasticNetPropensityModel()\n",
    "p = p_model.fit_predict(X, treatment)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:18.075149Z",
     "start_time": "2021-06-05T00:41:10.045499Z"
    },
    "nterop": {
     "id": "14"
    }
   },
   "outputs": [],
   "source": [
    "s_learner = BaseSRegressor(LGBMRegressor())\n",
    "s_ate = s_learner.estimate_ate(X, treatment, y)[0]\n",
    "s_ite = s_learner.fit_predict(X, treatment, y)\n",
    "\n",
    "t_learner = BaseTRegressor(LGBMRegressor())\n",
    "t_ate = t_learner.estimate_ate(X, treatment, y)[0][0]\n",
    "t_ite = t_learner.fit_predict(X, treatment, y)\n",
    "\n",
    "x_learner = BaseXRegressor(LGBMRegressor())\n",
    "x_ate = x_learner.estimate_ate(X, treatment, y, p)[0][0]\n",
    "x_ite = x_learner.fit_predict(X, treatment, y, p)\n",
    "\n",
    "r_learner = BaseRRegressor(LGBMRegressor())\n",
    "r_ate = r_learner.estimate_ate(X, treatment, y, p)[0][0]\n",
    "r_ite = r_learner.fit_predict(X, treatment, y, p)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:31.390411Z",
     "start_time": "2021-06-05T00:41:18.077142Z"
    },
    "nterop": {
     "id": "15"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "10/10 [==============================] - 5s 156ms/step - loss: 1790.3492 - regression_loss: 864.6742 - binary_classification_loss: 41.3394 - treatment_accuracy: 0.7299 - track_epsilon: 0.0063 - val_loss: 242.1589 - val_regression_loss: 87.0011 - val_binary_classification_loss: 32.6806 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0055\n",
      "Epoch 2/30\n",
      "10/10 [==============================] - 0s 7ms/step - loss: 311.9302 - regression_loss: 135.2392 - binary_classification_loss: 32.8420 - treatment_accuracy: 0.8643 - track_epsilon: 0.0059 - val_loss: 230.2209 - val_regression_loss: 79.8030 - val_binary_classification_loss: 34.3533 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0047\n",
      "Epoch 3/30\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 274.1216 - regression_loss: 118.1561 - binary_classification_loss: 31.3200 - treatment_accuracy: 0.8169 - track_epsilon: 0.0044 - val_loss: 238.4452 - val_regression_loss: 82.0530 - val_binary_classification_loss: 36.2376 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0049\n",
      "Epoch 4/30\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 205.4690 - regression_loss: 85.9585 - binary_classification_loss: 27.2440 - treatment_accuracy: 0.8606 - track_epsilon: 0.0053 - val_loss: 235.7122 - val_regression_loss: 78.5524 - val_binary_classification_loss: 39.7929 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0057\n",
      "Epoch 1/300\n",
      "10/10 [==============================] - 1s 41ms/step - loss: 195.6840 - regression_loss: 80.7820 - binary_classification_loss: 27.7316 - treatment_accuracy: 0.8497 - track_epsilon: 0.0054 - val_loss: 207.3960 - val_regression_loss: 67.6306 - val_binary_classification_loss: 38.1122 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0177\n",
      "Epoch 2/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 183.9956 - regression_loss: 75.3269 - binary_classification_loss: 26.4330 - treatment_accuracy: 0.8622 - track_epsilon: 0.0182 - val_loss: 197.0267 - val_regression_loss: 64.0559 - val_binary_classification_loss: 38.4298 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0117\n",
      "Epoch 3/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 178.8321 - regression_loss: 72.7892 - binary_classification_loss: 26.7587 - treatment_accuracy: 0.8555 - track_epsilon: 0.0081 - val_loss: 195.6257 - val_regression_loss: 63.5609 - val_binary_classification_loss: 38.2400 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0073\n",
      "Epoch 4/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 177.0419 - regression_loss: 71.8475 - binary_classification_loss: 27.1255 - treatment_accuracy: 0.8521 - track_epsilon: 0.0091 - val_loss: 200.6521 - val_regression_loss: 65.3493 - val_binary_classification_loss: 37.6216 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0082\n",
      "Epoch 5/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 198.0597 - regression_loss: 82.4320 - binary_classification_loss: 27.0536 - treatment_accuracy: 0.8497 - track_epsilon: 0.0076 - val_loss: 194.4365 - val_regression_loss: 63.1230 - val_binary_classification_loss: 37.8598 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0064\n",
      "Epoch 6/300\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 174.1273 - regression_loss: 70.2306 - binary_classification_loss: 27.7639 - treatment_accuracy: 0.8460 - track_epsilon: 0.0075 - val_loss: 194.3751 - val_regression_loss: 63.1176 - val_binary_classification_loss: 37.9318 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0100\n",
      "Epoch 7/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 187.2528 - regression_loss: 77.2338 - binary_classification_loss: 26.6574 - treatment_accuracy: 0.8545 - track_epsilon: 0.0094 - val_loss: 193.4222 - val_regression_loss: 62.8618 - val_binary_classification_loss: 37.8932 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0100\n",
      "Epoch 8/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 179.5122 - regression_loss: 72.3961 - binary_classification_loss: 28.5867 - treatment_accuracy: 0.8357 - track_epsilon: 0.0110 - val_loss: 196.3768 - val_regression_loss: 63.7690 - val_binary_classification_loss: 37.5827 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0094\n",
      "Epoch 9/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 180.4453 - regression_loss: 74.0497 - binary_classification_loss: 26.0765 - treatment_accuracy: 0.8582 - track_epsilon: 0.0077 - val_loss: 192.4576 - val_regression_loss: 62.1838 - val_binary_classification_loss: 37.8295 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0104\n",
      "Epoch 10/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 159.2664 - regression_loss: 63.3608 - binary_classification_loss: 26.5497 - treatment_accuracy: 0.8544 - track_epsilon: 0.0118 - val_loss: 192.8072 - val_regression_loss: 62.6150 - val_binary_classification_loss: 38.0268 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0147\n",
      "Epoch 11/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 157.3967 - regression_loss: 62.1042 - binary_classification_loss: 27.1641 - treatment_accuracy: 0.8496 - track_epsilon: 0.0139 - val_loss: 190.8307 - val_regression_loss: 61.5484 - val_binary_classification_loss: 37.7637 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0105\n",
      "Epoch 12/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 171.3408 - regression_loss: 69.0045 - binary_classification_loss: 27.2531 - treatment_accuracy: 0.8468 - track_epsilon: 0.0092 - val_loss: 190.0314 - val_regression_loss: 61.2129 - val_binary_classification_loss: 37.7777 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0087\n",
      "Epoch 13/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 162.0215 - regression_loss: 62.9260 - binary_classification_loss: 30.3296 - treatment_accuracy: 0.8189 - track_epsilon: 0.0084 - val_loss: 189.2025 - val_regression_loss: 60.8970 - val_binary_classification_loss: 37.7255 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0126\n",
      "Epoch 14/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 173.8554 - regression_loss: 70.6637 - binary_classification_loss: 26.2371 - treatment_accuracy: 0.8540 - track_epsilon: 0.0126 - val_loss: 189.1865 - val_regression_loss: 60.8833 - val_binary_classification_loss: 37.6686 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0099\n",
      "Epoch 15/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 157.4061 - regression_loss: 63.5324 - binary_classification_loss: 24.4340 - treatment_accuracy: 0.8718 - track_epsilon: 0.0093 - val_loss: 186.8761 - val_regression_loss: 60.0529 - val_binary_classification_loss: 37.9590 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0089\n",
      "Epoch 16/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 172.2053 - regression_loss: 69.6182 - binary_classification_loss: 26.8678 - treatment_accuracy: 0.8502 - track_epsilon: 0.0080 - val_loss: 188.4488 - val_regression_loss: 60.4575 - val_binary_classification_loss: 37.6228 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0113\n",
      "Epoch 17/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 162.3663 - regression_loss: 65.2318 - binary_classification_loss: 26.0364 - treatment_accuracy: 0.8562 - track_epsilon: 0.0105 - val_loss: 186.5810 - val_regression_loss: 59.7591 - val_binary_classification_loss: 37.6693 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0114\n",
      "Epoch 18/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 163.8466 - regression_loss: 65.5130 - binary_classification_loss: 26.7643 - treatment_accuracy: 0.8503 - track_epsilon: 0.0119 - val_loss: 190.8825 - val_regression_loss: 62.1830 - val_binary_classification_loss: 37.9718 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0087\n",
      "Epoch 19/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 167.6180 - regression_loss: 68.2214 - binary_classification_loss: 25.3027 - treatment_accuracy: 0.8620 - track_epsilon: 0.0066 - val_loss: 185.6225 - val_regression_loss: 59.4746 - val_binary_classification_loss: 37.6837 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0060\n",
      "Epoch 20/300\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10/10 [==============================] - 0s 6ms/step - loss: 168.5476 - regression_loss: 68.3371 - binary_classification_loss: 25.8652 - treatment_accuracy: 0.8578 - track_epsilon: 0.0079 - val_loss: 184.5200 - val_regression_loss: 59.2031 - val_binary_classification_loss: 37.5099 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0098\n",
      "Epoch 21/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 157.9161 - regression_loss: 63.3173 - binary_classification_loss: 25.2899 - treatment_accuracy: 0.8634 - track_epsilon: 0.0083 - val_loss: 185.0839 - val_regression_loss: 59.1452 - val_binary_classification_loss: 37.5366 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0079\n",
      "Epoch 22/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 160.4739 - regression_loss: 63.1629 - binary_classification_loss: 28.0595 - treatment_accuracy: 0.8358 - track_epsilon: 0.0086 - val_loss: 183.9351 - val_regression_loss: 59.1525 - val_binary_classification_loss: 37.5067 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0066\n",
      "Epoch 23/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 155.0890 - regression_loss: 60.5116 - binary_classification_loss: 28.0962 - treatment_accuracy: 0.8349 - track_epsilon: 0.0046 - val_loss: 183.6170 - val_regression_loss: 58.7653 - val_binary_classification_loss: 37.2800 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0051\n",
      "Epoch 24/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 149.4288 - regression_loss: 58.8520 - binary_classification_loss: 25.9568 - treatment_accuracy: 0.8546 - track_epsilon: 0.0052 - val_loss: 188.7191 - val_regression_loss: 60.7916 - val_binary_classification_loss: 37.1127 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0069\n",
      "Epoch 25/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 156.3095 - regression_loss: 61.0641 - binary_classification_loss: 28.1708 - treatment_accuracy: 0.8315 - track_epsilon: 0.0080 - val_loss: 182.5451 - val_regression_loss: 58.6875 - val_binary_classification_loss: 37.3179 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0075\n",
      "Epoch 26/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 154.8900 - regression_loss: 61.1673 - binary_classification_loss: 26.2975 - treatment_accuracy: 0.8542 - track_epsilon: 0.0059 - val_loss: 184.6580 - val_regression_loss: 59.6472 - val_binary_classification_loss: 37.4789 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0040\n",
      "Epoch 27/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 153.7275 - regression_loss: 60.6342 - binary_classification_loss: 26.5628 - treatment_accuracy: 0.8494 - track_epsilon: 0.0054 - val_loss: 187.6736 - val_regression_loss: 60.4940 - val_binary_classification_loss: 37.1448 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0046\n",
      "Epoch 28/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 159.4707 - regression_loss: 63.5524 - binary_classification_loss: 26.3749 - treatment_accuracy: 0.8515 - track_epsilon: 0.0043 - val_loss: 185.6613 - val_regression_loss: 60.3003 - val_binary_classification_loss: 37.4100 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0041\n",
      "Epoch 29/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 144.6116 - regression_loss: 57.2770 - binary_classification_loss: 24.2153 - treatment_accuracy: 0.8699 - track_epsilon: 0.0037 - val_loss: 183.7683 - val_regression_loss: 58.8389 - val_binary_classification_loss: 37.2161 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0046\n",
      "Epoch 30/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 156.1744 - regression_loss: 61.9692 - binary_classification_loss: 26.0622 - treatment_accuracy: 0.8516 - track_epsilon: 0.0058 - val_loss: 181.6741 - val_regression_loss: 58.3027 - val_binary_classification_loss: 37.2483 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0056\n",
      "Epoch 31/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 149.5090 - regression_loss: 59.6090 - binary_classification_loss: 24.2077 - treatment_accuracy: 0.8685 - track_epsilon: 0.0052 - val_loss: 183.1471 - val_regression_loss: 58.8850 - val_binary_classification_loss: 37.3616 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0048\n",
      "Epoch 32/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 158.8317 - regression_loss: 63.1364 - binary_classification_loss: 26.3766 - treatment_accuracy: 0.8524 - track_epsilon: 0.0040 - val_loss: 182.8958 - val_regression_loss: 58.5878 - val_binary_classification_loss: 37.0665 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0047\n",
      "Epoch 33/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 153.5294 - regression_loss: 59.8976 - binary_classification_loss: 27.7524 - treatment_accuracy: 0.8380 - track_epsilon: 0.0065 - val_loss: 184.9241 - val_regression_loss: 59.4232 - val_binary_classification_loss: 36.8858 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0053\n",
      "Epoch 34/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 149.8718 - regression_loss: 59.2581 - binary_classification_loss: 25.4000 - treatment_accuracy: 0.8586 - track_epsilon: 0.0050 - val_loss: 181.1425 - val_regression_loss: 58.2419 - val_binary_classification_loss: 37.1219 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0023\n",
      "Epoch 35/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 147.3198 - regression_loss: 58.8689 - binary_classification_loss: 23.9842 - treatment_accuracy: 0.8717 - track_epsilon: 0.0020 - val_loss: 182.3788 - val_regression_loss: 58.5675 - val_binary_classification_loss: 37.1934 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0054\n",
      "Epoch 36/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 143.2286 - regression_loss: 55.5299 - binary_classification_loss: 26.2675 - treatment_accuracy: 0.8521 - track_epsilon: 0.0059 - val_loss: 183.0977 - val_regression_loss: 59.1656 - val_binary_classification_loss: 37.0963 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0033\n",
      "Epoch 37/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 149.7070 - regression_loss: 59.5447 - binary_classification_loss: 24.7418 - treatment_accuracy: 0.8625 - track_epsilon: 0.0023 - val_loss: 183.3780 - val_regression_loss: 58.6777 - val_binary_classification_loss: 37.0991 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0018\n",
      "Epoch 38/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 156.6596 - regression_loss: 62.0057 - binary_classification_loss: 26.7843 - treatment_accuracy: 0.8461 - track_epsilon: 0.0034 - val_loss: 183.2619 - val_regression_loss: 58.6840 - val_binary_classification_loss: 36.8925 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0030\n",
      "Epoch 39/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 155.1980 - regression_loss: 60.9731 - binary_classification_loss: 27.2335 - treatment_accuracy: 0.8443 - track_epsilon: 0.0027 - val_loss: 181.5153 - val_regression_loss: 58.3543 - val_binary_classification_loss: 37.0235 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0026\n",
      "Epoch 40/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 148.6864 - regression_loss: 58.6236 - binary_classification_loss: 25.4435 - treatment_accuracy: 0.8563 - track_epsilon: 0.0023 - val_loss: 181.4140 - val_regression_loss: 58.1816 - val_binary_classification_loss: 36.9654 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0040\n",
      "\n",
      "Epoch 00040: ReduceLROnPlateau reducing learning rate to 4.999999873689376e-06.\n",
      "Epoch 41/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 144.7339 - regression_loss: 56.2116 - binary_classification_loss: 26.5446 - treatment_accuracy: 0.8501 - track_epsilon: 0.0048 - val_loss: 180.4425 - val_regression_loss: 57.9971 - val_binary_classification_loss: 36.8370 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0046\n",
      "Epoch 42/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 152.6819 - regression_loss: 59.9780 - binary_classification_loss: 26.7807 - treatment_accuracy: 0.8460 - track_epsilon: 0.0035 - val_loss: 181.0754 - val_regression_loss: 58.0677 - val_binary_classification_loss: 36.8611 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0022\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 43/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 141.4455 - regression_loss: 55.1271 - binary_classification_loss: 25.5353 - treatment_accuracy: 0.8533 - track_epsilon: 0.0021 - val_loss: 181.6914 - val_regression_loss: 58.2652 - val_binary_classification_loss: 36.9244 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0029\n",
      "Epoch 44/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 143.5718 - regression_loss: 54.9356 - binary_classification_loss: 27.7278 - treatment_accuracy: 0.8368 - track_epsilon: 0.0035 - val_loss: 182.9616 - val_regression_loss: 58.7127 - val_binary_classification_loss: 36.7605 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0030\n",
      "Epoch 45/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 148.0318 - regression_loss: 58.2297 - binary_classification_loss: 25.4171 - treatment_accuracy: 0.8623 - track_epsilon: 0.0028 - val_loss: 182.2657 - val_regression_loss: 58.7829 - val_binary_classification_loss: 36.9492 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0031\n",
      "Epoch 46/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 148.2912 - regression_loss: 58.4527 - binary_classification_loss: 25.4111 - treatment_accuracy: 0.8550 - track_epsilon: 0.0034 - val_loss: 181.4299 - val_regression_loss: 58.1758 - val_binary_classification_loss: 36.8661 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0024\n",
      "Epoch 47/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 149.8918 - regression_loss: 57.9189 - binary_classification_loss: 28.0705 - treatment_accuracy: 0.8318 - track_epsilon: 0.0019 - val_loss: 181.7516 - val_regression_loss: 58.2234 - val_binary_classification_loss: 36.8086 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0017\n",
      "Epoch 48/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 143.9826 - regression_loss: 56.0217 - binary_classification_loss: 25.9636 - treatment_accuracy: 0.8518 - track_epsilon: 0.0018 - val_loss: 183.2366 - val_regression_loss: 58.7730 - val_binary_classification_loss: 36.7622 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0020\n",
      "\n",
      "Epoch 00048: ReduceLROnPlateau reducing learning rate to 2.499999936844688e-06.\n",
      "Epoch 49/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 141.9178 - regression_loss: 54.9617 - binary_classification_loss: 26.0167 - treatment_accuracy: 0.8551 - track_epsilon: 0.0025 - val_loss: 181.6692 - val_regression_loss: 58.2019 - val_binary_classification_loss: 36.8163 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0035\n",
      "Epoch 50/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 154.0470 - regression_loss: 60.6821 - binary_classification_loss: 26.7084 - treatment_accuracy: 0.8442 - track_epsilon: 0.0037 - val_loss: 181.5136 - val_regression_loss: 58.1967 - val_binary_classification_loss: 36.7926 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0030\n",
      "Epoch 51/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 154.2879 - regression_loss: 61.1156 - binary_classification_loss: 25.9554 - treatment_accuracy: 0.8530 - track_epsilon: 0.0026 - val_loss: 181.1187 - val_regression_loss: 58.0944 - val_binary_classification_loss: 36.7854 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0019\n",
      "Epoch 52/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 147.1910 - regression_loss: 57.9212 - binary_classification_loss: 25.5444 - treatment_accuracy: 0.8585 - track_epsilon: 0.0019 - val_loss: 180.9492 - val_regression_loss: 58.0477 - val_binary_classification_loss: 36.8212 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0023\n",
      "Epoch 53/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 144.5095 - regression_loss: 57.1991 - binary_classification_loss: 24.5623 - treatment_accuracy: 0.8633 - track_epsilon: 0.0025 - val_loss: 181.2697 - val_regression_loss: 58.0844 - val_binary_classification_loss: 36.8072 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0025\n",
      "Epoch 54/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 149.1749 - regression_loss: 58.6545 - binary_classification_loss: 26.4255 - treatment_accuracy: 0.8508 - track_epsilon: 0.0027 - val_loss: 181.6855 - val_regression_loss: 58.2050 - val_binary_classification_loss: 36.8246 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0024\n",
      "Epoch 55/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 143.3488 - regression_loss: 57.4108 - binary_classification_loss: 22.5247 - treatment_accuracy: 0.8810 - track_epsilon: 0.0018 - val_loss: 181.5240 - val_regression_loss: 58.1889 - val_binary_classification_loss: 36.8670 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0017\n",
      "Epoch 56/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 145.3050 - regression_loss: 57.0797 - binary_classification_loss: 25.3895 - treatment_accuracy: 0.8567 - track_epsilon: 0.0020 - val_loss: 181.2868 - val_regression_loss: 58.1061 - val_binary_classification_loss: 36.7694 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0020\n",
      "Epoch 57/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 150.1354 - regression_loss: 58.8804 - binary_classification_loss: 26.4138 - treatment_accuracy: 0.8482 - track_epsilon: 0.0023 - val_loss: 181.1185 - val_regression_loss: 58.1035 - val_binary_classification_loss: 36.8033 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0027\n",
      "Epoch 58/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 145.2017 - regression_loss: 57.0173 - binary_classification_loss: 25.5667 - treatment_accuracy: 0.8535 - track_epsilon: 0.0028 - val_loss: 181.2872 - val_regression_loss: 58.0891 - val_binary_classification_loss: 36.7888 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0023\n",
      "Epoch 59/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 147.7799 - regression_loss: 57.4570 - binary_classification_loss: 26.7010 - treatment_accuracy: 0.8456 - track_epsilon: 0.0022 - val_loss: 181.2169 - val_regression_loss: 58.0942 - val_binary_classification_loss: 36.7608 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0024\n",
      "\n",
      "Epoch 00059: ReduceLROnPlateau reducing learning rate to 1.249999968422344e-06.\n",
      "Epoch 60/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 143.9625 - regression_loss: 56.2486 - binary_classification_loss: 25.4459 - treatment_accuracy: 0.8584 - track_epsilon: 0.0021 - val_loss: 180.9821 - val_regression_loss: 58.0493 - val_binary_classification_loss: 36.7809 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0018\n",
      "Epoch 61/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 145.4550 - regression_loss: 55.9254 - binary_classification_loss: 27.6142 - treatment_accuracy: 0.8356 - track_epsilon: 0.0018 - val_loss: 181.0581 - val_regression_loss: 58.0568 - val_binary_classification_loss: 36.7708 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0018\n",
      "Epoch 62/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 146.6129 - regression_loss: 57.8608 - binary_classification_loss: 24.9736 - treatment_accuracy: 0.8610 - track_epsilon: 0.0019 - val_loss: 181.3456 - val_regression_loss: 58.1192 - val_binary_classification_loss: 36.7725 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0019\n",
      "Epoch 63/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 148.2220 - regression_loss: 58.4927 - binary_classification_loss: 25.3438 - treatment_accuracy: 0.8560 - track_epsilon: 0.0019 - val_loss: 181.4519 - val_regression_loss: 58.1659 - val_binary_classification_loss: 36.7662 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0019\n",
      "Epoch 64/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 140.0457 - regression_loss: 53.7192 - binary_classification_loss: 26.7202 - treatment_accuracy: 0.8449 - track_epsilon: 0.0018 - val_loss: 181.4884 - val_regression_loss: 58.1761 - val_binary_classification_loss: 36.7904 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0018\n",
      "\n",
      "Epoch 00064: ReduceLROnPlateau reducing learning rate to 6.24999984211172e-07.\n",
      "Epoch 65/300\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10/10 [==============================] - 0s 6ms/step - loss: 138.7515 - regression_loss: 54.9897 - binary_classification_loss: 22.9222 - treatment_accuracy: 0.8764 - track_epsilon: 0.0019 - val_loss: 181.4227 - val_regression_loss: 58.1461 - val_binary_classification_loss: 36.7599 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0021\n",
      "Epoch 66/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 150.2262 - regression_loss: 59.0825 - binary_classification_loss: 26.3098 - treatment_accuracy: 0.8488 - track_epsilon: 0.0021 - val_loss: 181.3573 - val_regression_loss: 58.1316 - val_binary_classification_loss: 36.7465 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0022\n",
      "Epoch 67/300\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 149.6757 - regression_loss: 59.4637 - binary_classification_loss: 24.7313 - treatment_accuracy: 0.8603 - track_epsilon: 0.0021 - val_loss: 181.3223 - val_regression_loss: 58.1212 - val_binary_classification_loss: 36.7481 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0021\n",
      "Epoch 68/300\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 151.1728 - regression_loss: 60.0713 - binary_classification_loss: 25.0154 - treatment_accuracy: 0.8615 - track_epsilon: 0.0021 - val_loss: 181.1846 - val_regression_loss: 58.0904 - val_binary_classification_loss: 36.7660 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0020\n",
      "Epoch 69/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 148.2535 - regression_loss: 59.0331 - binary_classification_loss: 24.1994 - treatment_accuracy: 0.8658 - track_epsilon: 0.0021 - val_loss: 181.2948 - val_regression_loss: 58.1039 - val_binary_classification_loss: 36.7439 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0023\n",
      "\n",
      "Epoch 00069: ReduceLROnPlateau reducing learning rate to 3.12499992105586e-07.\n",
      "Epoch 70/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 145.3151 - regression_loss: 57.0691 - binary_classification_loss: 25.4983 - treatment_accuracy: 0.8584 - track_epsilon: 0.0023 - val_loss: 181.3544 - val_regression_loss: 58.1245 - val_binary_classification_loss: 36.7449 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0023\n",
      "Epoch 71/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 150.1866 - regression_loss: 57.9752 - binary_classification_loss: 28.1966 - treatment_accuracy: 0.8297 - track_epsilon: 0.0023 - val_loss: 181.2958 - val_regression_loss: 58.1128 - val_binary_classification_loss: 36.7442 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0022\n",
      "Epoch 72/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 144.9820 - regression_loss: 57.4340 - binary_classification_loss: 24.4047 - treatment_accuracy: 0.8619 - track_epsilon: 0.0022 - val_loss: 181.3424 - val_regression_loss: 58.1227 - val_binary_classification_loss: 36.7440 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0023\n",
      "Epoch 73/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 148.8112 - regression_loss: 58.0125 - binary_classification_loss: 26.8692 - treatment_accuracy: 0.8447 - track_epsilon: 0.0022 - val_loss: 181.3199 - val_regression_loss: 58.1187 - val_binary_classification_loss: 36.7438 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0021\n",
      "Epoch 74/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 144.3984 - regression_loss: 56.9031 - binary_classification_loss: 24.6382 - treatment_accuracy: 0.8624 - track_epsilon: 0.0021 - val_loss: 181.3810 - val_regression_loss: 58.1361 - val_binary_classification_loss: 36.7440 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0021\n",
      "\n",
      "Epoch 00074: ReduceLROnPlateau reducing learning rate to 1.56249996052793e-07.\n",
      "Epoch 75/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 147.5547 - regression_loss: 57.7667 - binary_classification_loss: 26.1622 - treatment_accuracy: 0.8478 - track_epsilon: 0.0021 - val_loss: 181.3161 - val_regression_loss: 58.1183 - val_binary_classification_loss: 36.7473 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0021\n",
      "Epoch 76/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 140.5001 - regression_loss: 53.5784 - binary_classification_loss: 27.3214 - treatment_accuracy: 0.8388 - track_epsilon: 0.0021 - val_loss: 181.2723 - val_regression_loss: 58.1086 - val_binary_classification_loss: 36.7488 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0022\n",
      "Epoch 77/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 143.8736 - regression_loss: 55.9839 - binary_classification_loss: 26.1250 - treatment_accuracy: 0.8466 - track_epsilon: 0.0022 - val_loss: 181.2639 - val_regression_loss: 58.1073 - val_binary_classification_loss: 36.7513 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0021\n",
      "Epoch 78/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 146.6917 - regression_loss: 58.5758 - binary_classification_loss: 23.5315 - treatment_accuracy: 0.8700 - track_epsilon: 0.0022 - val_loss: 181.2961 - val_regression_loss: 58.1147 - val_binary_classification_loss: 36.7518 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0022\n",
      "Epoch 79/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 143.4007 - regression_loss: 54.8006 - binary_classification_loss: 27.7054 - treatment_accuracy: 0.8383 - track_epsilon: 0.0021 - val_loss: 181.3115 - val_regression_loss: 58.1188 - val_binary_classification_loss: 36.7477 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0021\n",
      "\n",
      "Epoch 00079: ReduceLROnPlateau reducing learning rate to 7.81249980263965e-08.\n",
      "Epoch 80/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 145.6183 - regression_loss: 57.3271 - binary_classification_loss: 25.1687 - treatment_accuracy: 0.8574 - track_epsilon: 0.0021 - val_loss: 181.2945 - val_regression_loss: 58.1152 - val_binary_classification_loss: 36.7475 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0021\n",
      "Epoch 81/300\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 142.3395 - regression_loss: 54.9129 - binary_classification_loss: 26.6164 - treatment_accuracy: 0.8461 - track_epsilon: 0.0021 - val_loss: 181.3037 - val_regression_loss: 58.1177 - val_binary_classification_loss: 36.7449 - val_treatment_accuracy: 0.7244 - val_track_epsilon: 0.0022\n"
     ]
    }
   ],
   "source": [
    "dragon = DragonNet(neurons_per_layer=200, targeted_reg=True)\n",
    "dragon_ite = dragon.fit_predict(X, treatment, y, return_components=False)\n",
    "dragon_ate = dragon_ite.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:31.631490Z",
     "start_time": "2021-06-05T00:41:31.391928Z"
    },
    "nterop": {
     "id": "16"
    }
   },
   "outputs": [],
   "source": [
    "df_preds = pd.DataFrame([s_ite.ravel(),\n",
    "                          t_ite.ravel(),\n",
    "                          x_ite.ravel(),\n",
    "                          r_ite.ravel(),\n",
    "                          dragon_ite.ravel(),\n",
    "                          tau.ravel(),\n",
    "                          treatment.ravel(),\n",
    "                          y.ravel()],\n",
    "                       index=['S','T','X','R','dragonnet','tau','w','y']).T\n",
    "\n",
    "df_cumgain = get_cumgain(df_preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:32.099368Z",
     "start_time": "2021-06-05T00:41:31.632920Z"
    },
    "nterop": {
     "id": "17"
    }
   },
   "outputs": [],
   "source": [
    "df_result = pd.DataFrame([s_ate, t_ate, x_ate, r_ate, dragon_ate, tau.mean()],\n",
    "                     index=['S','T','X','R','dragonnet','actual'], columns=['ATE'])\n",
    "df_result['MAE'] = [mean_absolute_error(t,p) for t,p in zip([s_ite, t_ite, x_ite, r_ite, dragon_ite],\n",
    "                                                            [tau.values.reshape(-1,1)]*5 )\n",
    "                ] + [None]\n",
    "df_result['AUUC'] = auuc_score(df_preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:32.225561Z",
     "start_time": "2021-06-05T00:41:32.100925Z"
    },
    "nterop": {
     "id": "18"
    }
   },
   "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>ATE</th>\n",
       "      <th>MAE</th>\n",
       "      <th>AUUC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>S</th>\n",
       "      <td>4.054511</td>\n",
       "      <td>1.027666</td>\n",
       "      <td>0.575822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>T</th>\n",
       "      <td>4.100199</td>\n",
       "      <td>0.980788</td>\n",
       "      <td>0.580929</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>X</th>\n",
       "      <td>4.020918</td>\n",
       "      <td>1.116303</td>\n",
       "      <td>0.564651</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R</th>\n",
       "      <td>4.257976</td>\n",
       "      <td>1.665557</td>\n",
       "      <td>0.556855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dragonnet</th>\n",
       "      <td>4.006536</td>\n",
       "      <td>1.165061</td>\n",
       "      <td>0.556426</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>actual</th>\n",
       "      <td>4.098887</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                ATE       MAE      AUUC\n",
       "S          4.054511  1.027666  0.575822\n",
       "T          4.100199  0.980788  0.580929\n",
       "X          4.020918  1.116303  0.564651\n",
       "R          4.257976  1.665557  0.556855\n",
       "dragonnet  4.006536  1.165061  0.556426\n",
       "actual     4.098887       NaN       NaN"
      ]
     },
     "execution_count": 14,
     "metadata": {
      "nterop": {
       "id": "45"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:41:33.085932Z",
     "start_time": "2021-06-05T00:41:32.226978Z"
    },
    "nterop": {
     "id": "20"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "nterop": {
       "id": "46"
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_gain(df_preds)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {
    "nterop": {
     "id": "22"
    }
   },
   "source": [
    "## Synthetic Data Generation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:48:12.467012Z",
     "start_time": "2021-06-05T00:41:33.184802Z"
    },
    "nterop": {
     "id": "23"
    }
   },
   "outputs": [],
   "source": [
    "y, X, w, tau, b, e = simulate_nuisance_and_easy_treatment(n=1000)\n",
    "\n",
    "X_train, X_val, y_train, y_val, w_train, w_val, tau_train, tau_val, b_train, b_val, e_train, e_val = \\\n",
    "    train_test_split(X, y, w, tau, b, e, test_size=0.2, random_state=123, shuffle=True)\n",
    "\n",
    "preds_dict_train = {}\n",
    "preds_dict_valid = {}\n",
    "\n",
    "preds_dict_train['Actuals'] = tau_train\n",
    "preds_dict_valid['Actuals'] = tau_val\n",
    "\n",
    "preds_dict_train['generated_data'] = {\n",
    "    'y': y_train,\n",
    "    'X': X_train,\n",
    "    'w': w_train,\n",
    "    'tau': tau_train,\n",
    "    'b': b_train,\n",
    "    'e': e_train}\n",
    "preds_dict_valid['generated_data'] = {\n",
    "    'y': y_val,\n",
    "    'X': X_val,\n",
    "    'w': w_val,\n",
    "    'tau': tau_val,\n",
    "    'b': b_val,\n",
    "    'e': e_val}\n",
    "\n",
    "# Predict p_hat because e would not be directly observed in real-life\n",
    "p_model = ElasticNetPropensityModel()\n",
    "p_hat_train = p_model.fit_predict(X_train, w_train)\n",
    "p_hat_val = p_model.fit_predict(X_val, w_val)\n",
    "\n",
    "for base_learner, label_l in zip([BaseSRegressor, BaseTRegressor, BaseXRegressor, BaseRRegressor],\n",
    "                                 ['S', 'T', 'X', 'R']):\n",
    "    for model, label_m in zip([LinearRegression, XGBRegressor], ['LR', 'XGB']):\n",
    "        # RLearner will need to fit on the p_hat\n",
    "        if label_l != 'R':\n",
    "            learner = base_learner(model())\n",
    "            # fit the model on training data only\n",
    "            learner.fit(X=X_train, treatment=w_train, y=y_train)\n",
    "            try:\n",
    "                preds_dict_train['{} Learner ({})'.format(\n",
    "                    label_l, label_m)] = learner.predict(X=X_train, p=p_hat_train).flatten()\n",
    "                preds_dict_valid['{} Learner ({})'.format(\n",
    "                    label_l, label_m)] = learner.predict(X=X_val, p=p_hat_val).flatten()\n",
    "            except TypeError:\n",
    "                preds_dict_train['{} Learner ({})'.format(\n",
    "                    label_l, label_m)] = learner.predict(X=X_train, treatment=w_train, y=y_train).flatten()\n",
    "                preds_dict_valid['{} Learner ({})'.format(\n",
    "                    label_l, label_m)] = learner.predict(X=X_val, treatment=w_val, y=y_val).flatten()\n",
    "        else:\n",
    "            learner = base_learner(model())\n",
    "            learner.fit(X=X_train, p=p_hat_train, treatment=w_train, y=y_train)\n",
    "            preds_dict_train['{} Learner ({})'.format(\n",
    "                label_l, label_m)] = learner.predict(X=X_train).flatten()\n",
    "            preds_dict_valid['{} Learner ({})'.format(\n",
    "                label_l, label_m)] = learner.predict(X=X_val).flatten()\n",
    "\n",
    "learner = DragonNet(verbose=False)\n",
    "learner.fit(X_train, treatment=w_train, y=y_train)\n",
    "preds_dict_train['DragonNet'] = learner.predict_tau(X=X_train).flatten()\n",
    "preds_dict_valid['DragonNet'] = learner.predict_tau(X=X_val).flatten()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:48:12.711611Z",
     "start_time": "2021-06-05T00:48:12.468459Z"
    },
    "nterop": {
     "id": "24"
    }
   },
   "outputs": [],
   "source": [
    "actuals_train = preds_dict_train['Actuals']\n",
    "actuals_validation = preds_dict_valid['Actuals']\n",
    "\n",
    "synthetic_summary_train = pd.DataFrame({label: [preds.mean(), mse(preds, actuals_train)] for label, preds\n",
    "                                        in preds_dict_train.items() if 'generated' not in label.lower()},\n",
    "                                       index=['ATE', 'MSE']).T\n",
    "synthetic_summary_train['Abs % Error of ATE'] = np.abs(\n",
    "    (synthetic_summary_train['ATE']/synthetic_summary_train.loc['Actuals', 'ATE']) - 1)\n",
    "\n",
    "synthetic_summary_validation = pd.DataFrame({label: [preds.mean(), mse(preds, actuals_validation)]\n",
    "                                             for label, preds in preds_dict_valid.items()\n",
    "                                             if 'generated' not in label.lower()},\n",
    "                                            index=['ATE', 'MSE']).T\n",
    "synthetic_summary_validation['Abs % Error of ATE'] = np.abs(\n",
    "    (synthetic_summary_validation['ATE']/synthetic_summary_validation.loc['Actuals', 'ATE']) - 1)\n",
    "\n",
    "# calculate kl divergence for training\n",
    "for label in synthetic_summary_train.index:\n",
    "    stacked_values = np.hstack((preds_dict_train[label], actuals_train))\n",
    "    stacked_low = np.percentile(stacked_values, 0.1)\n",
    "    stacked_high = np.percentile(stacked_values, 99.9)\n",
    "    bins = np.linspace(stacked_low, stacked_high, 100)\n",
    "\n",
    "    distr = np.histogram(preds_dict_train[label], bins=bins)[0]\n",
    "    distr = np.clip(distr/distr.sum(), 0.001, 0.999)\n",
    "    true_distr = np.histogram(actuals_train, bins=bins)[0]\n",
    "    true_distr = np.clip(true_distr/true_distr.sum(), 0.001, 0.999)\n",
    "\n",
    "    kl = entropy(distr, true_distr)\n",
    "    synthetic_summary_train.loc[label, 'KL Divergence'] = kl\n",
    "\n",
    "# calculate kl divergence for validation\n",
    "for label in synthetic_summary_validation.index:\n",
    "    stacked_values = np.hstack((preds_dict_valid[label], actuals_validation))\n",
    "    stacked_low = np.percentile(stacked_values, 0.1)\n",
    "    stacked_high = np.percentile(stacked_values, 99.9)\n",
    "    bins = np.linspace(stacked_low, stacked_high, 100)\n",
    "\n",
    "    distr = np.histogram(preds_dict_valid[label], bins=bins)[0]\n",
    "    distr = np.clip(distr/distr.sum(), 0.001, 0.999)\n",
    "    true_distr = np.histogram(actuals_validation, bins=bins)[0]\n",
    "    true_distr = np.clip(true_distr/true_distr.sum(), 0.001, 0.999)\n",
    "\n",
    "    kl = entropy(distr, true_distr)\n",
    "    synthetic_summary_validation.loc[label, 'KL Divergence'] = kl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:48:13.372709Z",
     "start_time": "2021-06-05T00:48:12.712843Z"
    },
    "nterop": {
     "id": "25"
    }
   },
   "outputs": [],
   "source": [
    "df_preds_train = pd.DataFrame([preds_dict_train['S Learner (LR)'].ravel(),\n",
    "                               preds_dict_train['S Learner (XGB)'].ravel(),\n",
    "                               preds_dict_train['T Learner (LR)'].ravel(),\n",
    "                               preds_dict_train['T Learner (XGB)'].ravel(),\n",
    "                               preds_dict_train['X Learner (LR)'].ravel(),\n",
    "                               preds_dict_train['X Learner (XGB)'].ravel(),\n",
    "                               preds_dict_train['R Learner (LR)'].ravel(),\n",
    "                               preds_dict_train['R Learner (XGB)'].ravel(),                               \n",
    "                               preds_dict_train['DragonNet'].ravel(),\n",
    "                               preds_dict_train['generated_data']['tau'].ravel(),\n",
    "                               preds_dict_train['generated_data']['w'].ravel(),\n",
    "                               preds_dict_train['generated_data']['y'].ravel()],\n",
    "                              index=['S Learner (LR)','S Learner (XGB)',\n",
    "                                     'T Learner (LR)','T Learner (XGB)',\n",
    "                                     'X Learner (LR)','X Learner (XGB)',\n",
    "                                     'R Learner (LR)','R Learner (XGB)',\n",
    "                                     'DragonNet','tau','w','y']).T\n",
    "\n",
    "synthetic_summary_train['AUUC'] = auuc_score(df_preds_train).iloc[:-1]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:48:13.591008Z",
     "start_time": "2021-06-05T00:48:13.374045Z"
    },
    "nterop": {
     "id": "26"
    }
   },
   "outputs": [],
   "source": [
    "df_preds_validation = pd.DataFrame([preds_dict_valid['S Learner (LR)'].ravel(),\n",
    "                               preds_dict_valid['S Learner (XGB)'].ravel(),\n",
    "                               preds_dict_valid['T Learner (LR)'].ravel(),\n",
    "                               preds_dict_valid['T Learner (XGB)'].ravel(),\n",
    "                               preds_dict_valid['X Learner (LR)'].ravel(),\n",
    "                               preds_dict_valid['X Learner (XGB)'].ravel(),\n",
    "                               preds_dict_valid['R Learner (LR)'].ravel(),\n",
    "                               preds_dict_valid['R Learner (XGB)'].ravel(),                               \n",
    "                               preds_dict_valid['DragonNet'].ravel(),\n",
    "                               preds_dict_valid['generated_data']['tau'].ravel(),\n",
    "                               preds_dict_valid['generated_data']['w'].ravel(),\n",
    "                               preds_dict_valid['generated_data']['y'].ravel()],\n",
    "                              index=['S Learner (LR)','S Learner (XGB)',\n",
    "                                     'T Learner (LR)','T Learner (XGB)',\n",
    "                                     'X Learner (LR)','X Learner (XGB)',\n",
    "                                     'R Learner (LR)','R Learner (XGB)',\n",
    "                                     'DragonNet','tau','w','y']).T\n",
    "\n",
    "synthetic_summary_validation['AUUC'] = auuc_score(df_preds_validation).iloc[:-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:48:13.764418Z",
     "start_time": "2021-06-05T00:48:13.592189Z"
    },
    "nterop": {
     "id": "27"
    }
   },
   "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>ATE</th>\n",
       "      <th>MSE</th>\n",
       "      <th>Abs % Error of ATE</th>\n",
       "      <th>KL Divergence</th>\n",
       "      <th>AUUC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Actuals</th>\n",
       "      <td>0.484486</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S Learner (LR)</th>\n",
       "      <td>0.528743</td>\n",
       "      <td>0.044194</td>\n",
       "      <td>0.091349</td>\n",
       "      <td>3.473087</td>\n",
       "      <td>0.508067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S Learner (XGB)</th>\n",
       "      <td>0.358208</td>\n",
       "      <td>0.310652</td>\n",
       "      <td>0.260643</td>\n",
       "      <td>0.817620</td>\n",
       "      <td>0.544115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>T Learner (LR)</th>\n",
       "      <td>0.493815</td>\n",
       "      <td>0.022688</td>\n",
       "      <td>0.019255</td>\n",
       "      <td>0.289978</td>\n",
       "      <td>0.610855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>T Learner (XGB)</th>\n",
       "      <td>0.397053</td>\n",
       "      <td>1.350928</td>\n",
       "      <td>0.180466</td>\n",
       "      <td>1.452143</td>\n",
       "      <td>0.521719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>X Learner (LR)</th>\n",
       "      <td>0.493815</td>\n",
       "      <td>0.022688</td>\n",
       "      <td>0.019255</td>\n",
       "      <td>0.289978</td>\n",
       "      <td>0.610855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>X Learner (XGB)</th>\n",
       "      <td>0.341352</td>\n",
       "      <td>0.620992</td>\n",
       "      <td>0.295435</td>\n",
       "      <td>1.086086</td>\n",
       "      <td>0.534827</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R Learner (LR)</th>\n",
       "      <td>0.457692</td>\n",
       "      <td>0.028116</td>\n",
       "      <td>0.055304</td>\n",
       "      <td>0.335083</td>\n",
       "      <td>0.614414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R Learner (XGB)</th>\n",
       "      <td>0.434709</td>\n",
       "      <td>4.575591</td>\n",
       "      <td>0.102741</td>\n",
       "      <td>1.907325</td>\n",
       "      <td>0.505088</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DragonNet</th>\n",
       "      <td>0.410899</td>\n",
       "      <td>0.044120</td>\n",
       "      <td>0.151888</td>\n",
       "      <td>0.467829</td>\n",
       "      <td>0.611620</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      ATE       MSE  Abs % Error of ATE  KL Divergence  \\\n",
       "Actuals          0.484486  0.000000            0.000000       0.000000   \n",
       "S Learner (LR)   0.528743  0.044194            0.091349       3.473087   \n",
       "S Learner (XGB)  0.358208  0.310652            0.260643       0.817620   \n",
       "T Learner (LR)   0.493815  0.022688            0.019255       0.289978   \n",
       "T Learner (XGB)  0.397053  1.350928            0.180466       1.452143   \n",
       "X Learner (LR)   0.493815  0.022688            0.019255       0.289978   \n",
       "X Learner (XGB)  0.341352  0.620992            0.295435       1.086086   \n",
       "R Learner (LR)   0.457692  0.028116            0.055304       0.335083   \n",
       "R Learner (XGB)  0.434709  4.575591            0.102741       1.907325   \n",
       "DragonNet        0.410899  0.044120            0.151888       0.467829   \n",
       "\n",
       "                     AUUC  \n",
       "Actuals               NaN  \n",
       "S Learner (LR)   0.508067  \n",
       "S Learner (XGB)  0.544115  \n",
       "T Learner (LR)   0.610855  \n",
       "T Learner (XGB)  0.521719  \n",
       "X Learner (LR)   0.610855  \n",
       "X Learner (XGB)  0.534827  \n",
       "R Learner (LR)   0.614414  \n",
       "R Learner (XGB)  0.505088  \n",
       "DragonNet        0.611620  "
      ]
     },
     "execution_count": 20,
     "metadata": {
      "nterop": {
       "id": "47"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "synthetic_summary_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:48:13.912363Z",
     "start_time": "2021-06-05T00:48:13.765680Z"
    },
    "nterop": {
     "id": "29"
    }
   },
   "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>ATE</th>\n",
       "      <th>MSE</th>\n",
       "      <th>Abs % Error of ATE</th>\n",
       "      <th>KL Divergence</th>\n",
       "      <th>AUUC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Actuals</th>\n",
       "      <td>0.511242</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S Learner (LR)</th>\n",
       "      <td>0.528743</td>\n",
       "      <td>0.042236</td>\n",
       "      <td>0.034233</td>\n",
       "      <td>4.574498</td>\n",
       "      <td>0.495423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S Learner (XGB)</th>\n",
       "      <td>0.434208</td>\n",
       "      <td>0.260496</td>\n",
       "      <td>0.150680</td>\n",
       "      <td>0.854890</td>\n",
       "      <td>0.544206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>T Learner (LR)</th>\n",
       "      <td>0.541503</td>\n",
       "      <td>0.025840</td>\n",
       "      <td>0.059191</td>\n",
       "      <td>0.686602</td>\n",
       "      <td>0.604712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>T Learner (XGB)</th>\n",
       "      <td>0.483404</td>\n",
       "      <td>0.679398</td>\n",
       "      <td>0.054452</td>\n",
       "      <td>1.215394</td>\n",
       "      <td>0.526918</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>X Learner (LR)</th>\n",
       "      <td>0.541503</td>\n",
       "      <td>0.025840</td>\n",
       "      <td>0.059191</td>\n",
       "      <td>0.686602</td>\n",
       "      <td>0.604712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>X Learner (XGB)</th>\n",
       "      <td>0.330427</td>\n",
       "      <td>0.344865</td>\n",
       "      <td>0.353678</td>\n",
       "      <td>1.227041</td>\n",
       "      <td>0.536599</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R Learner (LR)</th>\n",
       "      <td>0.510236</td>\n",
       "      <td>0.030801</td>\n",
       "      <td>0.001967</td>\n",
       "      <td>0.654228</td>\n",
       "      <td>0.608133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R Learner (XGB)</th>\n",
       "      <td>0.417823</td>\n",
       "      <td>1.990451</td>\n",
       "      <td>0.182730</td>\n",
       "      <td>1.650560</td>\n",
       "      <td>0.504991</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DragonNet</th>\n",
       "      <td>0.462146</td>\n",
       "      <td>0.043679</td>\n",
       "      <td>0.096032</td>\n",
       "      <td>0.825673</td>\n",
       "      <td>0.605744</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      ATE       MSE  Abs % Error of ATE  KL Divergence  \\\n",
       "Actuals          0.511242  0.000000            0.000000       0.000000   \n",
       "S Learner (LR)   0.528743  0.042236            0.034233       4.574498   \n",
       "S Learner (XGB)  0.434208  0.260496            0.150680       0.854890   \n",
       "T Learner (LR)   0.541503  0.025840            0.059191       0.686602   \n",
       "T Learner (XGB)  0.483404  0.679398            0.054452       1.215394   \n",
       "X Learner (LR)   0.541503  0.025840            0.059191       0.686602   \n",
       "X Learner (XGB)  0.330427  0.344865            0.353678       1.227041   \n",
       "R Learner (LR)   0.510236  0.030801            0.001967       0.654228   \n",
       "R Learner (XGB)  0.417823  1.990451            0.182730       1.650560   \n",
       "DragonNet        0.462146  0.043679            0.096032       0.825673   \n",
       "\n",
       "                     AUUC  \n",
       "Actuals               NaN  \n",
       "S Learner (LR)   0.495423  \n",
       "S Learner (XGB)  0.544206  \n",
       "T Learner (LR)   0.604712  \n",
       "T Learner (XGB)  0.526918  \n",
       "X Learner (LR)   0.604712  \n",
       "X Learner (XGB)  0.536599  \n",
       "R Learner (LR)   0.608133  \n",
       "R Learner (XGB)  0.504991  \n",
       "DragonNet        0.605744  "
      ]
     },
     "execution_count": 21,
     "metadata": {
      "nterop": {
       "id": "48"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "synthetic_summary_validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:48:14.332740Z",
     "start_time": "2021-06-05T00:48:13.913521Z"
    },
    "nterop": {
     "id": "31"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "nterop": {
       "id": "49"
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_gain(df_preds_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-05T00:48:14.785782Z",
     "start_time": "2021-06-05T00:48:14.335017Z"
    },
    "nterop": {
     "id": "33"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "nterop": {
       "id": "50"
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_gain(df_preds_validation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "nterop": {
     "id": "35"
    }
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "hide_input": false,
  "kernelspec": {
   "display_name": "causalml-py37",
   "language": "python",
   "name": "causalml-py37"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.10"
  },
  "nterop": {
   "seedId": "50"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {
    "height": "174px",
    "width": "252px"
   },
   "number_sections": false,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "165px"
   },
   "toc_section_display": "block",
   "toc_window_display": true
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}