Chapter_07.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from scipy import stats\n",
    "\n",
    "import dowhy\n",
    "from dowhy import CausalModel\n",
    "\n",
    "from sklearn.linear_model import LassoCV\n",
    "from sklearn.ensemble import GradientBoostingRegressor\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "plt.style.use('fivethirtyeight')\n",
    "\n",
    "import graphviz"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.8'"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dowhy.__version__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "COLORS = [\n",
    "    '#00B0F0',\n",
    "    '#FF0000'\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Chapter 07\n",
    "\n",
    "In this chapter we focus on the full four-steps causal inference process. First, we introduce DoWhy and EconML libraries. Next, we discuss and exercise how to encode causal information in a form of a graph that is understandable to DoWhy. Next, we demonstrate how to find applicable estimands for a problem at hand. Then, we compute causal effect estimates using DoWhy and estimators provided in EconML. To understand how good our estimates are, we perform a series of refutal tests. Finally, we run a full (and more complex) causal process using DoWhy and EconML."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Causal models with DoWhy & EconML"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Generate the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# First, we'll build a structural causal model (SCM)\n",
    "class GPSMemorySCM:\n",
    "    \n",
    "    def __init__(self, random_seed=None):\n",
    "        self.random_seed = random_seed\n",
    "        self.u_x = stats.truncnorm(0, np.infty, scale=5)\n",
    "        self.u_y = stats.norm(scale=2)\n",
    "        self.u_z = stats.norm(scale=2)\n",
    "        self.u = stats.truncnorm(0, np.infty, scale=4)\n",
    "        \n",
    "    def sample(self, sample_size=100, treatment_value=None):\n",
    "        \"\"\"Samples from the SCM\"\"\"\n",
    "        if self.random_seed:\n",
    "            np.random.seed(self.random_seed)\n",
    "        \n",
    "        u_x = self.u_x.rvs(sample_size)\n",
    "        u_y = self.u_y.rvs(sample_size)\n",
    "        u_z = self.u_z.rvs(sample_size)\n",
    "        u = self.u.rvs(sample_size)\n",
    "        \n",
    "        if treatment_value:\n",
    "            gps = np.array([treatment_value]*sample_size)\n",
    "        else:\n",
    "            gps = u_x + 0.7*u\n",
    "            \n",
    "        hippocampus = -0.6*gps + 0.25*u_z\n",
    "        memory = 0.7*hippocampus + 0.25*u\n",
    "        \n",
    "        return gps, hippocampus, memory\n",
    "    \n",
    "    def intervene(self, treatment_value, sample_size=100):\n",
    "        \"\"\"Intervenes on the SCM\"\"\"\n",
    "        return self.sample(treatment_value=treatment_value, sample_size=sample_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Instantiate the SCM\n",
    "scm = GPSMemorySCM()\n",
    "\n",
    "# Generate observational data\n",
    "gps_obs, hippocampus_obs, memory_obs = scm.sample(1000)\n",
    "\n",
    "# Encode as a pandas df\n",
    "df = pd.DataFrame(np.vstack([gps_obs, hippocampus_obs, memory_obs]).T, columns=['X', 'Z', 'Y'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>X</th>\n",
       "      <th>Z</th>\n",
       "      <th>Y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4.913517</td>\n",
       "      <td>-3.336259</td>\n",
       "      <td>-1.062827</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6.554867</td>\n",
       "      <td>-4.153368</td>\n",
       "      <td>-1.067321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8.884924</td>\n",
       "      <td>-4.715010</td>\n",
       "      <td>-1.892951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>13.985812</td>\n",
       "      <td>-9.348827</td>\n",
       "      <td>-5.768484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.552113</td>\n",
       "      <td>-1.295731</td>\n",
       "      <td>-0.365418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>4.449338</td>\n",
       "      <td>-2.199691</td>\n",
       "      <td>-0.586440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>5.257915</td>\n",
       "      <td>-2.793845</td>\n",
       "      <td>-1.281048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>0.729923</td>\n",
       "      <td>-0.789898</td>\n",
       "      <td>-0.320348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>2.311561</td>\n",
       "      <td>-0.406512</td>\n",
       "      <td>-0.272665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>17.640751</td>\n",
       "      <td>-10.177374</td>\n",
       "      <td>-5.517893</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             X          Z         Y\n",
       "0     4.913517  -3.336259 -1.062827\n",
       "1     6.554867  -4.153368 -1.067321\n",
       "2     8.884924  -4.715010 -1.892951\n",
       "3    13.985812  -9.348827 -5.768484\n",
       "4     1.552113  -1.295731 -0.365418\n",
       "..         ...        ...       ...\n",
       "995   4.449338  -2.199691 -0.586440\n",
       "996   5.257915  -2.793845 -1.281048\n",
       "997   0.729923  -0.789898 -0.320348\n",
       "998   2.311561  -0.406512 -0.272665\n",
       "999  17.640751 -10.177374 -5.517893\n",
       "\n",
       "[1000 rows x 3 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 1: Modeling the problem"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Step 1.1 - Define the graph - `GML`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create the graph describing the causal structure\n",
    "gml_graph = \"\"\"\n",
    "graph [\n",
    "    directed 1\n",
    "    \n",
    "    node [\n",
    "        id \"X\" \n",
    "        label \"X\"\n",
    "    ]    \n",
    "    node [\n",
    "        id \"Z\"\n",
    "        label \"Z\"\n",
    "    ]\n",
    "    node [\n",
    "        id \"Y\"\n",
    "        label \"Y\"\n",
    "    ]\n",
    "    node [\n",
    "        id \"U\"\n",
    "        label \"U\"\n",
    "    ]\n",
    "    \n",
    "    edge [\n",
    "        source \"X\"\n",
    "        target \"Z\"\n",
    "    ]\n",
    "    edge [\n",
    "        source \"Z\"\n",
    "        target \"Y\"\n",
    "    ]\n",
    "    edge [\n",
    "        source \"U\"\n",
    "        target \"X\"\n",
    "    ]\n",
    "    edge [\n",
    "        source \"U\"\n",
    "        target \"Y\"\n",
    "    ]\n",
    "]\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Step 1.2 - define the DoWhy model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# With graph\n",
    "model = CausalModel(\n",
    "    data=df,\n",
    "    treatment='X',\n",
    "    outcome='Y',\n",
    "    graph=gml_graph\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.view_model()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 2: Identify the estimand"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Estimand type: nonparametric-ate\n",
      "\n",
      "### Estimand : 1\n",
      "Estimand name: backdoor\n",
      "No such variable(s) found!\n",
      "\n",
      "### Estimand : 2\n",
      "Estimand name: iv\n",
      "No such variable(s) found!\n",
      "\n",
      "### Estimand : 3\n",
      "Estimand name: frontdoor\n",
      "Estimand expression:\n",
      " ⎡ d       d       ⎤\n",
      "E⎢────(Y)⋅────([Z])⎥\n",
      " ⎣d[Z]    d[X]     ⎦\n",
      "Estimand assumption 1, Full-mediation: Z intercepts (blocks) all directed paths from X to Y.\n",
      "Estimand assumption 2, First-stage-unconfoundedness: If U→{X} and U→{Z} then P(Z|X,U) = P(Z|X)\n",
      "Estimand assumption 3, Second-stage-unconfoundedness: If U→{Z} and U→Y then P(Y|Z, X, U) = P(Y|Z, X)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "estimand = model.identify_effect()\n",
    "print(estimand)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 3: Estimate the causal effect"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example 1 - Linear Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "two_stage_regression\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': None, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': None, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "Estimate of causal effect (linear regression): -0.4101007729201311\n"
     ]
    }
   ],
   "source": [
    "estimate = model.estimate_effect(\n",
    "    identified_estimand=estimand,\n",
    "    method_name='frontdoor.two_stage_regression')\n",
    "\n",
    "print(f'Estimate of causal effect (linear regression): {estimate.value}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 4: Run refutation tests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': False, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': [], 'first_stage_model': None, 'second_stage_model': None}\n"
     ]
    }
   ],
   "source": [
    "refute_subset = model.refute_estimate(\n",
    "    estimand=estimand, \n",
    "    estimate=estimate,\n",
    "    method_name=\"data_subset_refuter\", \n",
    "    subset_fraction=0.4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Refute: Use a subset of data\n",
      "Estimated effect:-0.4101007729201311\n",
      "New effect:-0.4094252079526228\n",
      "p value:0.98\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(refute_subset)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Full example"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n",
       " \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n",
       "<!-- Generated by graphviz version 2.46.0 (20210118.1747)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"206pt\" height=\"224pt\"\n",
       " viewBox=\"0.00 0.00 206.00 224.00\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g id=\"graph0\" class=\"graph\" transform=\"scale(1 1) rotate(0) translate(4 220)\">\n",
       "<polygon fill=\"white\" stroke=\"transparent\" points=\"-4,4 -4,-220 202,-220 202,4 -4,4\"/>\n",
       "<!-- S -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>S</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"99\" cy=\"-198\" rx=\"27\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"99\" y=\"-194.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">S</text>\n",
       "</g>\n",
       "<!-- Q -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Q</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"171\" cy=\"-90\" rx=\"27\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"171\" y=\"-86.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">Q</text>\n",
       "</g>\n",
       "<!-- S&#45;&gt;Q -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>S&#45;&gt;Q</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M110.25,-181.13C122.03,-163.45 140.67,-135.49 154.2,-115.2\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"157.19,-117.02 159.83,-106.76 151.37,-113.14 157.19,-117.02\"/>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-18\" rx=\"27\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-14.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- S&#45;&gt;Y -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>S&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M91.97,-180.41C79.35,-148.88 52.81,-82.52 37.85,-45.12\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"40.92,-43.38 33.96,-35.39 34.42,-45.98 40.92,-43.38\"/>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"171\" cy=\"-18\" rx=\"27\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"171\" y=\"-14.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Q&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>Q&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M171,-71.83C171,-64.13 171,-54.97 171,-46.42\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"174.5,-46.41 171,-36.41 167.5,-46.41 174.5,-46.41\"/>\n",
       "</g>\n",
       "<!-- Q&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>Q&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M149.13,-79.06C124.78,-66.89 85.24,-47.12 57.77,-33.39\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"59.11,-30.14 48.6,-28.8 55.98,-36.4 59.11,-30.14\"/>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge7\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M143.91,-18C121.53,-18 89.5,-18 64.63,-18\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"64.45,-14.5 54.45,-18 64.45,-21.5 64.45,-14.5\"/>\n",
       "</g>\n",
       "<!-- P -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>P</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-162\" rx=\"27\" ry=\"18\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-158.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">P</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;P -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>X&#45;&gt;P</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M155.86,-33.14C130.54,-58.46 79.46,-109.54 49.78,-139.22\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"46.97,-137.09 42.37,-146.63 51.91,-142.03 46.97,-137.09\"/>\n",
       "</g>\n",
       "<!-- Y&#45;&gt;P -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>Y&#45;&gt;P</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M27,-36.24C27,-60.8 27,-104.75 27,-133.65\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"23.5,-133.91 27,-143.91 30.5,-133.91 23.5,-133.91\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x26f8e00d5b0>"
      ]
     },
     "execution_count": 248,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "graph = graphviz.Digraph(format='png', engine='neato')\n",
    "\n",
    "nodes = ['S', 'Q', 'X', 'Y', 'P']\n",
    "positions = ['2,2.5!', '3,1!', '3,0!', '1, 0!', '1,2!', '1.7,0.9!']\n",
    "\n",
    "edges = ['SQ', 'SY', 'QX', 'QY', 'XP', 'YP', 'XY']\n",
    "\n",
    "[graph.node(n, pos=pos) for n, pos in zip(nodes, positions)]\n",
    "graph.edges(edges)\n",
    "\n",
    "graph.render(f'img/ch_07_full_example')\n",
    "\n",
    "graph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 301,
   "metadata": {},
   "outputs": [],
   "source": [
    "SAMPLE_SIZE = 1000\n",
    "\n",
    "S = np.random.random(SAMPLE_SIZE)\n",
    "Q = 0.2*S + 0.67*np.random.random(SAMPLE_SIZE)\n",
    "X = 0.14*Q + 0.4*np.random.random(SAMPLE_SIZE)\n",
    "Y = 0.7*X + 0.11*Q + 0.32*S + 0.24*np.random.random(SAMPLE_SIZE)\n",
    "P = 0.43*X + 0.21*Y + 0.22*np.random.random(SAMPLE_SIZE)\n",
    "\n",
    "# Encode as a pandas df\n",
    "df = pd.DataFrame(np.vstack([S, Q, X, Y, P]).T, columns=['S', 'Q', 'X', 'Y', 'P'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Generate the GML graph\n",
    "gml_string = 'graph [directed 1\\n'\n",
    "\n",
    "for node in nodes:\n",
    "    gml_string += f'\\tnode [id \"{node}\" label \"{node}\"]\\n'\n",
    "\n",
    "for edge in edges:\n",
    "    gml_string += f'\\tedge [source \"{edge[0]}\" target \"{edge[1]}\"]\\n'\n",
    "    \n",
    "gml_string += ']'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 303,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "graph [directed 1\n",
      "\tnode [id \"S\" label \"S\"]\n",
      "\tnode [id \"Q\" label \"Q\"]\n",
      "\tnode [id \"X\" label \"X\"]\n",
      "\tnode [id \"Y\" label \"Y\"]\n",
      "\tnode [id \"P\" label \"P\"]\n",
      "\tedge [source \"S\" target \"Q\"]\n",
      "\tedge [source \"S\" target \"Y\"]\n",
      "\tedge [source \"Q\" target \"X\"]\n",
      "\tedge [source \"Q\" target \"Y\"]\n",
      "\tedge [source \"X\" target \"P\"]\n",
      "\tedge [source \"Y\" target \"P\"]\n",
      "\tedge [source \"X\" target \"Y\"]\n",
      "]\n"
     ]
    }
   ],
   "source": [
    "print(gml_string)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 291,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Instantiate the CausalModel \n",
    "model = CausalModel(\n",
    "    data=df,\n",
    "    treatment='X',\n",
    "    outcome='Y',\n",
    "    graph=gml_string\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.view_model()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Estimand type: nonparametric-ate\n",
      "\n",
      "### Estimand : 1\n",
      "Estimand name: backdoor\n",
      "Estimand expression:\n",
      " d          \n",
      "────(E[Y|Q])\n",
      "d[X]        \n",
      "Estimand assumption 1, Unconfoundedness: If U→{X} and U→Y then P(Y|X,Q,U) = P(Y|X,Q)\n",
      "\n",
      "### Estimand : 2\n",
      "Estimand name: iv\n",
      "No such variable(s) found!\n",
      "\n",
      "### Estimand : 3\n",
      "Estimand name: frontdoor\n",
      "No such variable(s) found!\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Get the estimand\n",
    "estimand = model.identify_effect()\n",
    "print(estimand)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 318,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Estimate of causal effect (DML): 0.6981455013596706\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n"
     ]
    }
   ],
   "source": [
    "estimate = model.estimate_effect(\n",
    "    identified_estimand=estimand,\n",
    "    method_name='backdoor.econml.dml.DML',\n",
    "    method_params={\n",
    "        'init_params': {\n",
    "            'model_y': GradientBoostingRegressor(),\n",
    "            'model_t': GradientBoostingRegressor(),\n",
    "            'model_final': LassoCV(fit_intercept=False),\n",
    "        },\n",
    "        'fit_params': {}}\n",
    ")\n",
    "\n",
    "print(f'Estimate of causal effect (DML): {estimate.value}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 316,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "linear_regression\n",
      "{'control_value': 0, 'treatment_value': 1, 'test_significance': None, 'evaluate_effect_strength': False, 'confidence_intervals': False, 'target_units': 'ate', 'effect_modifiers': []}\n",
      "Estimate of causal effect (linear regression): 0.688147600764658\n"
     ]
    }
   ],
   "source": [
    "estimate_lr = model.estimate_effect(\n",
    "    identified_estimand=estimand,\n",
    "    method_name='backdoor.linear_regression')\n",
    "\n",
    "print(f'Estimate of causal effect (linear regression): {estimate_lr.value}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 319,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n"
     ]
    }
   ],
   "source": [
    "random_cause = model.refute_estimate(\n",
    "    estimand=estimand, \n",
    "    estimate=estimate,\n",
    "    method_name='random_common_cause'\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 320,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Refute: Add a random common cause\n",
      "Estimated effect:0.6981455013596706\n",
      "New effect:0.6691543110272069\n",
      "p value:0.1399999999999999\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(random_cause)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 321,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n"
     ]
    }
   ],
   "source": [
    "placebo_refuter = model.refute_estimate(\n",
    "    estimand=estimand, \n",
    "    estimate=estimate,\n",
    "    method_name='placebo_treatment_refuter'\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 322,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Refute: Use a Placebo Treatment\n",
      "Estimated effect:0.6981455013596706\n",
      "New effect:0.0\n",
      "p value:2.0\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(placebo_refuter)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## GCM API"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 323,
   "metadata": {},
   "outputs": [],
   "source": [
    "import networkx as nx\n",
    "from dowhy import gcm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 411,
   "metadata": {},
   "outputs": [],
   "source": [
    "SAMPLE_SIZE = 1000\n",
    "\n",
    "S = np.random.random(SAMPLE_SIZE)\n",
    "Q = 0.2*S + 0.67*np.random.random(SAMPLE_SIZE)\n",
    "X = 0.14*Q + 0.4*np.random.random(SAMPLE_SIZE)\n",
    "Y = 0.7*X + 0.11*Q + 0.32*S + 0.24*np.random.random(SAMPLE_SIZE)\n",
    "P = 0.43*X + 0.21*Y + 0.22*np.random.random(SAMPLE_SIZE)\n",
    "\n",
    "# Encode as a pandas df\n",
    "df = pd.DataFrame(np.vstack([S, Q, X, Y, P]).T, columns=['S', 'Q', 'X', 'Y', 'P'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 413,
   "metadata": {},
   "outputs": [],
   "source": [
    "edges = ['SQ', 'SY', 'QX', 'QY', 'XP', 'YP', 'XY']\n",
    "graph_nx = nx.DiGraph([(edge[0], edge[1]) for edge in edges])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 473,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAb4AAAEuCAYAAADx63eqAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8/fFQqAAAACXBIWXMAAAsTAAALEwEAmpwYAAA9qUlEQVR4nO3deVxU9f4/8NeZhWGRTVGuW/zcyCU1NE1SRNnBBTV3xL2riFaWpvbVsrJ8lK3GuOVSbhWuCYqyqIhpamqmmeJyc0kTTPZ9Zs7vD4OrXpcZmJkzzLye/92Cc952kde8P+dz3h9BFEURRERENkImdQFERETmxOAjIiKbwuAjIiKbwuAjIiKbwuAjIiKbwuAjIiKbwuAjIiKbwuAjIiKbwuAjIiKbwuAjIiKbwuAjIiKbwuAjIiKbwuAjIiKbwuAjIiKbwuAjIiKbwuAjIiKbopC6gCfRiiKO52iRlKXBL3laFGhEaEVALgDOCgHPusoR3kCBzu5yyAVB6nKJiMjCCZZ6AntWmQ4fZJYhLVuDS0U6lOge/bWOMqBFHRkCPRSY461CAxUbWSIiejiLCz6tKOLt30ux7noFrpYYXtpTDgKimyjxbht7yNgBEhHRAywq+C4XaRF1vBjHc3WoqEFVSgHo7CbDhs6OaO4kN16BRERU61lM8P2ap8GQYyXILHrMmqaBvJ1k2NTFAR1cLf5RJhERmYlFBN/lIi3CDxcbNfQqeTvJkOTLzo+IiO6SfBeIVhQRddw0oQcAmUU6jDpeAq30+U5ERBZA8uB7+/dSHM81TehV+jlXi/nnykx6DyIiqh0kDb6sMh3WXa+o0UYWfVSIwLpr5cguM23AEhGR5ZM0+BZmllXrlYXquFIiYmEmuz4iIlsn2XZHrSgiNVuj99f/J9gZnioBWhEo0ojYlaXBtF9LUKTV/56p2RpoRZETXoiIbJhkHd/xHC0uGbihpd+RIjjvzEen9EJ0cZNj7tMqg77/UpEOJ3INSEoiIrI6kgVfUpbmsWPIHudGqYikWxo842zYKwrFOmD3Lf27TCIisj6SBd+p/Op3Xk3sBUR4KnAyz/BrVOd7iIjIekj2jC+/Gls5t3d1gkYE8ipE7LxVgQ+qsVmlQMP3+YiIbJmEm1sM/54BR4uQll2zjo25R0Rk2yRb6pRLtLFSwQ2dREQ2TbLgc1FKk0DOTD4iIpsmWfB1dJFmaLSPK4dVExHZMsmCL7yBAg4G3L1ZSkGNn+85yoAwTx5RRERkyyQ7lkgrinh2XyHOFJhvfmbdgpvom/YRyktLkJ2djZycHMyfPx/9+vUzWw1ERCQtydofuSAgqL4CZwrKzXZP8eQ+rP16TdX/ViqVcHd3N9v9iYhIepIOqZ7jrcJTDubZbOLlIODE3Gi0atUKMtndP7Yoijhy5AhycnLMUgMREUlP0uBroJIhuokSpt7gqRSA0U3t8P/c62D//v2oW7culEolXn/9dZw8eRLNmzfHxIkTcfLkSdMWQkREkpPsGV8lrSiiR0Yhfsox3bM+X3c5Dvo5QfbPqQynT5/GuHHjkJ6eDicnJ9y6dQsrV67EsmXL8NRTTyE2NhaDBw+GnZ2dyWoiIiJpSB58AHC5SIvww8XINPC0Bn24Fd9G0vOO6Na84RO/VqPRICEhAXFxcTh79iwmTpyISZMmoUmTJkavi4iIpCHpUmel5k5ybOriAG8n45bjpahA7ux+6OHdFG3btsXChQtx6dKlR369QqHAwIEDkZaWhr179yI3NxcdOnTA4MGDsW/fPljAZwQiIqohi+j4Kl0u0mLU8RL8nKtFNWZYV1EKwHNucmzo7ICwTm2RmZkJAJDL5WjYsCGuXbum97UKCgqwbt06qNVqAEBsbCyio6Ph7Oxc/QKJiEgyFtHxVWruJEeGnxNmtVLBq5q7Pb0cBMxqpcJBPyc0c5IjNjYWDg4OAO4G344dOwy6nrOzM6ZMmYIzZ85ArVZj79698PLywtSpU/H7779Xq0YiIpKORXV898oq02FhZhnSsjW4VKRD8WMe/znKgBZOMgTVV2COtwr1Vf/N81u3bqFp06ZQKpVo2LAhBg8ejIULF0IQqr+V9Pr161i+fDm++uortGvXDlOnTkW/fv2gUHAqDBGRpbPY4KukFUWcyNVi9y0NTuZpUaARoRHvnrLgrBDg4ypHmKcCndzkkD8izGbPno3+/fvj6aefRnBwMPz9/fHpp5/WKPwAoLy8HFu2bEFcXByuXbuGSZMm4aWXXkKDBg1qdF0iIjIdiw8+Y8vJyUFYWBg6d+6MuLi4qpfZa+qXX36BWq3G5s2b0adPH8TGxqJbt241DlciIjIumws+AMjPz0dERARat26N5cuXQy433okNOTk5WLNmDZYsWQJXV1fExsZixIgRVc8ZiYhIWjYZfABQWFiIfv36oWnTpli9erXRn8/pdDrs2bMHarUaR44cwdixYxETE4PmzZsb9T5ERGQYi9rVaU516tTBzp078ddff2HUqFGoqKgw6vVlMhnCw8ORmJiII0eOAACef/559OnTB0lJSdDpzHcqBRER/ZfNdnyVSktL8eKLL0KlUuG7774z6ZiykpISfPfdd4iLi0NeXh5iYmIwbtw41K1b12T3JCKi+9lsx1fJ3t4e27ZtgyiKGDRoEEpLS012LwcHB4wbNw4///wz1q9fj19++QUtWrTggGwiIjOy+eADADs7O8THx8PJyQmRkZEoLi426f0EQUC3bt2wbt06nD9/Hs2bN0dkZCS6d++OjRs3orzcfGcUEhHZGptf6ryXRqPBuHHj8OeffyIhIQFOTk5mvXdCQgLUajV+++03DsgmIjIRdnz3UCgU+Prrr9GsWTOEhYUhPz/frPceOHAgUlNTOSCbiMiE2PE9hE6nw9SpU3HixAns3r0bbm5uktRRUFCA9evXIy4uDgAHZBMRGQOD7xFEUcT06dORkZGB5ORk1KtXT9Ja0tPToVarkZaWhpEjRyI2NhZt2rSRrCYiotqKS52PIAgCPvvsMwQFBSEgIABZWVmS1tKrVy9s2rQJv/76K+rWrYuAgAAEBgZi69at0Gg0ktVGRFTbsON7AlEUMX/+fGzevBmpqalo2PDJJ7mbQ+WAbLVajatXr2LSpEmYOHEiPD09pS6NiMiiseN7AkEQ8M4772DkyJHw9/fH9evXpS4JwN1XMEaMGIGDBw9ix44d+OOPP9C6dWuMGjUKhw8f5mYYIqJHYMdngI8//hhLly6tOozW0tw7INvFxQVTp07F8OHD4ejoKHVpREQWg8FnoC+//BKffPIJ0tLS0KJFC6nLeSidTofk5GTExcXhyJEjGDNmDGJiYiy2XiIic+JSp4GmTZuGN998E7169cL58+elLuehZDIZwsLCqgZky2QydOvWDX369MGuXbs4IJuIbBo7vmpas2YN5s6di+TkZLRr107qcp6IA7KJiO5i8NXAxo0b8frrr2P37t3o2LGj1OXoRRRFHDlyBGq1GomJiXjxxRcRGxsLHx8fqUsjIjILBl8Nbd68GVOnTkViYiKee+45qcsxSFZWFlauXIlly5ahSZMmmDp1KgYPHmzSo5mIiKTG4DOCH374AS+99BJ27NiBbt26SV2OwTQaDRITExEXF4czZ85UDchu2rSp1KURERkdN7cYQWRkJL755hv0798fGRkZUpdjMIVCgQEDBiA1NRX79+9Hfn4+OnbsiBdffJEDsonI6rDjM6LU1FSMHDkS3333HQICAqQup0YqB2Sr1WqIoogpU6Zg9OjRHJBNRLUeg8/I0tPTMWTIEKxbtw6hoaFSl1NjHJBNRNaGS51G5u/vj+3btyM6OhoJCQlSl1NjHJBNRNaGHZ+JHDt2DH379sXSpUsxaNAgqcsxqvLycmzduhVxcXG4cuUKJk+ezAHZRFRrsOMzkS5dumDPnj2IjY3Fd999J3U5RmVnZ4fhw4fj4MGDSEhIwJUrV9C6dWtERUVxQDYRWTx2fCZ25swZhISEYOHChRgzZozU5ZhMTk4Ovv76a6jVari4uCA2NhYjRozggGwisjgMPjM4d+4cgoKCMH/+fEycOFHqckyqckC2Wq3GTz/9xAHZRGRxuNRpBq1bt8b+/fvx3nvvQa1WS12OSVUOyE5ISMDRo0c5IJuILA47PjP6448/EBAQgGnTpmH69OlSl2M2lQOy1Wo1cnJyMGXKFA7IJiLJMPjM7Nq1awgICMD48eMxZ84cqcsxK1EUcfToUcTFxXFANhFJhsEngRs3biAwMBDDhw/HW2+9BUEQpC7J7LKysrBq1SosXboUTZo0QWxsLAYPHgyVSiV1aURk5Rh8Erl16xaCgoLQr18/vP/++zYZfsB/B2Sr1WqcPn2aA7KJyOS4uUUinp6e2LdvH5KSkjBjxgybffetckB2SkpK1YDsZ599Fi+++CL27t1rs/9diMh02PFJLCcnB6GhoejatSsWL14MmYyfRe4dkK3T6RAbG4vo6Gi4uLhIXRoRWQEGnwXIy8tDREQE2rZti+XLlzP8/iGKIg4cOAC1Wo3U1FSMGDECsbGxaNu2rdSlEVEtxt+wFsDV1RV79uzBhQsXMG7cOGi1WqlLsgiCIMDf3x/x8fE4ffo0PDw8EBgYiICAAGzZsoUDsomoWtjxWZDi4mJERkbCw8MDa9euhVKplLoki1M5IFutVuOPP/7ApEmT8NJLL3FANhHpjR2fBXF0dERCQgLy8/MxfPhwlJeXS12SxakckJ2RkYHExERcvXq1akD2oUOHuBmGiJ6IHZ8FKisrw7Bhw6DT6bBp0ya+2/YElQOylyxZgjp16mDq1KkckE1Ej8Tgs1AVFRWIiopCfn4+tm3bBgcHB6lLsnj3Dsg+fPgwxowZgylTpnBANhHdh0udFkqpVGLjxo3w8PBAnz59UFRUJHVJFu/eAdnHjh2DQqFAt27dEBERgZ07d3JANhEBYMdn8bRaLV566SVcvHgRO3fuhLOzs9Ql1SolJSX4/vvvERcXh5ycHMTExGD8+PEckE1kw9jxWTi5XI6VK1eiXbt2CAkJQW5urtQl1SoODg4YO3Ysjh07ho0bN+LXX39FixYtMGHCBJw4cULq8ohIAgy+WkAmk2HJkiXo2rUrgoKCcOfOHalLqnUEQcDzzz+PtWvX4vz582jZsiUGDhwIX19fbNiwAWVlZVKXSERmwqXOWkQURcyaNQt79uxBamoq6tevL3VJtZpGo8HOnTsRFxeH06dPY8KECZg8eTIHZBNZOXZ8tYggCPjwww8RGRmJXr164a+//pK6pFpNoVAgMjISKSkpSE9PR2FhITp27IhBgwYhLS2N7wQSWSl2fLXUggULsH79eqSlpaFx48ZSl2M1CgsLqwZka7VaTJkyBaNHj+aAbCIrwuCrxRYtWoTly5cjLS0NXl5eUpdjVTggm8h6camzFps5cyZefvll9OrVC5cvX5a6HKvyqAHZvXv35oBsolqOHZ8VWL58Od5//32kpqbC29tb6nKsVnl5ObZt24a4uDj85z//qRqQ/a9//Uvq0ojIAOz4rMCkSZPwzjvvICAgAGfPnpW6HKtlZ2eHYcOGISMjAzt37sS1a9fQpk0bjBw5Ej/++CM3wxDVEuz4rMiGDRswc+ZM7N69Gx06dJC6HJuQm5uLr7/+Gmq1GnXq1EFsbCxGjhzJAdlEFozBZ2U2bdqEadOmYdeuXejUqZPU5dgMnU6HlJQUqNVqHDp0CGPGjEFMTAxatmwpdWlE9AAudVqZIUOGYPny5QgPD8eRI0ekLsdmyGQyhIaGYseOHVUDsn19fREeHo6dO3dCq9VKXSIR/YMdn5XatWsXxo4di61bt6JHjx5Sl2OTKgdkq9Vq/P3331UDsuvVqyd1aUQ2jR2flYqIiMCGDRswaNAg7Nu3T+pybNK9A7K/++47nD59Gi1btsT48eNx/Phxqcsjslns+Kxceno6hgwZgvXr1yMkJETqcmxednY2Vq5ciWXLlqFRo0aIjY3FkCFDoFKppC6NyGYw+GzAoUOHMGDAAKxZswZ9+vSRuhzC3XMWExMToVarcerUKUycOBGTJk3CU089JXVpRFaPS5024IUXXkBiYiLGjx+Pbdu2SV0O4e45i5GRkUhOTsaBAwdQWFgIHx8fDBw4kAOyiUyMHZ8NOXnyJCIiIvD5559j2LBhUpdDD7h3QLZGo8GUKVMwZswYDsgmMjIGn405ffo0QkND8eGHHyI6OlrqcughRFFERkYG4uLikJqaiuHDhyM2Nhbt2rWTujQiq8ClThvTvn17pKWlYc6cOVi1apXU5dBDCIKAnj17Ij4+HmfOnEGDBg0QHByMXr16YfPmzaioqJC6RKJajR2fjbpw4QKCgoIwa9YsTJkyRepy6Ak4IJvIeNjx2ahWrVph//79+Pjjj/H5559LXQ49wYMDsq9fv442bdpgxIgROHjwIDfDEBmAHZ+Nu3r1KgIDAzFx4kTMmjVL6nLIAJUDspcsWQJHR8eqAdlOTk5Sl0Zk0Rh8hBs3biAgIAAjR47EvHnzIAiC1CWRAR4ckD169GhMmTKFA7KJHoFLnYRGjRohPT0dmzZtwty5c7lsVss8OCBbqVTC19cXYWFhSExM5IBsogew46Mqt2/fRnBwMAIDA7Fo0SJ2frVYSUkJ4uPjERcXh9u3b2PKlCkckE30D3Z8VMXDwwNpaWlIT0/HK6+8ws6vFnNwcMCYMWNw7NgxfP/99zhz5gxatGiBcePG4eeff5a6PCJJseOj/5GXl4fw8HC0b98eS5cuhUzGz0fWIDs7G6tWrcLSpUvRsGFDxMbGYujQoRyQTTaHwUcPVVBQgL59+6JZs2ZYtWoV5HK51CWRkTw4IHvChAmYPHkyB2STzWDw0SMVFRUhMjISnp6e+Oabb6BQKKQuiYzs/PnzWLJkCdavXw8/Pz9MnToVgYGBZnm+qxVFHM/RIilLg1/ytCjQiNCKgFwAnBUCnnWVI7yBAp3d5ZDzeTMZEYOPHqukpASDBg2Ck5MTNm7cCDs7O6lLIhMoLCzEhg0bEBcXh4qKCsTGxmL06NFwdXU1+r2yynT4ILMMadkaXCrSoUT36K91lAEt6sgQ6KHAHG8VGqi47E41x+CjJyorK8PQoUMhiiI2bdrEZ0JWrHJAtlqtRnJyctWA7GeeeabG19aKIt7+vRTrrlfgaonhv3aechAQ3USJd9vYQ8YOkGqAwUd6qaiowMiRI1FYWIitW7fCwcFB6pLIxG7cuIEVK1ZgxYoV8Pb2RmxsLAYMGAClUmnwtS4XaRF1vBjHc3WoqMFvHKUAdHaTYUNnRzR34nNnqh4GH+lNo9FgzJgxuHXrFn744QeOxrIRlQOy1Wo1Ll26VDUgu2HDhnp9/695Ggw5VoLMosesaRrI20mGTV0c0MGVz53JcFwwJ70pFAqsXbsWTZs2RUREBAoKCqQuicygckD2gQMHkJSUhD///BNt27bF8OHDnzgg+3KR1uihBwCZRToMOVaCy0WcSkOGY8dHBtPpdIiJicHp06eRlJRkkg0QZNlyc3PxzTffQK1Ww8HBAbGxsYiKirpvFUAriuiRUYifcowbevfydZcjw8+Juz7JIOz4yGAymQzLli3Dc889h6CgINy5c0fqksjM3Nzc8Morr+DcuXNYtGgREhMT8dRTT2H69Om4cOECAODt30txPNd0oQcAP+dqMf9cmUnvQdaHwUfVIggCvvjiC/Tq1QuBgYG4ffu21CWRBGQyGUJCQrBjxw4cP34cKpUKL7zwAsKGjcK66xU12siijwoRWHetHNllpg1Ysi5c6qQaEUUR8+bNw/bt25GWlgZPT0+pSyKJlZaWInr/dWwuq2+2e05vbodP23OnMemHHR/ViCAIWLBgAYYNGwZ/f3/8+eefUpdEElOqVDhnp/8HoPWdHLDq2ftDq2c9OW6HO+NfKv2e3aVma6DlZ3jSE4OPjGLevHkYP348evXqhatXr0pdDknoeI4WlwzYxfny6VJEeCoQVP/uqwkqGfDVsw54/Uwp/irTL8wuFelwIpc7PEk/DD4ymjfeeAOxsbHw9/fHf/7zH6nLIYkkZWkeO4bsQXcqREw7XYIVHR3gKAfeflqFS0U6fHOtQu9rFOuA3bc01aiWbBHf/iSjevXVV6FSqeDv74+0tDS0atVK6pLIzE7lG955bb6hwbDGWnzb2RHd68nhs7/Q4GuczGPHR/ph8JHRxcTEQKVSoXfv3khJSUGbNm2kLonMKL+aWzljfy3BpSBn/N/vpbhWjVmeBRo+4yP9MPjIJMaPHw87OzsEBgZiz549aN++vdQlkZloq5k/WWUibpeL+K2gep0bc4/0xeAjkxk1ahSUSiWCg4Oxa9cudOrUSeqSyAzkEg1RUXB4C+mJwUcmNWzYMNjZ2SE8PBwJCQno2rWr1CWRibkopUkgZyYf6Ym7OsnkBg4ciNWrV6Nv37748ccfpS6HTKyjizTHBfm48pgi0g+Dj8yiT58+WL9+PQYOHIj9+/dLXQ6ZUHgDBRyq+ZulWUoB0rINf8bnKAPCPLmARfph8JHZhISEID4+HkOHDkVKSorU5ZCJdHaXo4WTeX+1tHCSoZMbOz7SD4OPzKpXr17Ytm0boqKisGvXLqnLoRrSarV4+umn4enpiUaNGqFBgwZwdXZGw+xMs9YRWF/Bo4lIbww+Mrvu3bsjISEB48aNw/bt26Uuh2pALpejfv36yMrKws2bN5GdnQ2lUokv/LzwlIN5gsjLQcCb3iqz3IusA4OPJPH8888jKSkJkydPRnx8vNTlUDWIooi0tDQUFxdD+KfbcnR0xMGDB9GmkQeimyhh6g2eSgEY3dQO9VX8VUb6408LSaZTp05ITk7Gq6++ivXr10tdDulJFEXs2rUL3bt3x5QpU/DKK6/Ax8cHcrkcy5YtQ7t27QAA77SxR2c30/6Kec5Njvmt2e2RYbgNiiTVoUMHpKamIiQkBOXl5Rg/frzUJdEj6HQ6bN++HQsWLEBFRQXmzp2LwYMHQy6Xo23btti9ezeio6Orvl4uCNjQ2RHhh4uRacBpDfpqjBKcHN0LMaH+GDFiBLp37w6lUmn0+5D14UG0ZBEyMzMRFBSEN998E5MnT5a6HLqHRqNBfHw83n//fTg6OmLu3Lno168fZDL9urlf8zQYcqzEqOHn7STD12118HuqHrRaLerUqQOtVosBAwZg48aNRrsPWScudZJF8Pb2xv79+/Hhhx/iiy++kLocAlBRUYHVq1ejTZs2WLJkCT799FMcPXoUkZGReoceAHRwVSDJ1xG+7vIaP/NTCoCvuxy7fR3h28gVQ4cOhSAIKCwshCiKcHDgKez0ZOz4yKJcuXIFgYGB+Pe//4033nhD6nJsUmlpKdasWYMPP/wQLVu2xNy5c+Hv71+1gaW6tKKI+efKsO5aOa5U4/QFLwcB0U3t8E5rFWT/1HLw4EGEh4ejrKwMMpkMP/30E5599tka1UnWj8FHFufPP/9EYGAgoqKiMG/ePKnLsRlFRUVYsWIFPv74Y/j4+OD//u//4Ovra/T7ZJXpsDCzDGnZGlwq0qH4MSugjrK7L6cH1Vdgjrfqf3ZviqKIJk2aoHnz5pg0aRJef/11bNu2DS+88ILR6ybrwc0tZHEaN26M/fv3IzAwEOXl5Xj33Xdr3G3Qo+Xn50OtVuPzzz+Hn58fEhMT4ePjY7L7NVDJ8Fl7B2hFESdytdh9S4P1h39FgQZ4uk1bKIS7A6d9XOUI81Sgk5v8kS+nC4KAQ4cOoUGDBnBwcEC9evUQGRmJ9evXIzQ01GR/Bqrd2PGRxcrOzkZwcDBCQkLw4YcfMvyM7M6dO1i8eDHUajVCQ0MxZ86cqlcRzOn27dto3LgxBEFAQUFBjXdm/vjjjxg0aBAWL16MYcOGGalKsibc3EIWq379+ti7dy/27duHV199FfyMZhxZWVmYPXs2WrVqhWvXruHQoUNYv369JKEHAK+88go0Gg0AIDExscbX6969O1JSUvDaa69h+fLlNb4eWR8GH1m0unXrIjU1FUePHkVMTAx0OuO/D2Yr/vzzT7z66qto3bo1CgoKcOLECaxatQqtWrWSrKZDhw5h27Zt0Ol0KCsrw8cff2yU63bo0AEHDhzAhx9+iIULF/JDE92HwUcWz9XVFcnJyTh79iwmTpwIrdbwY2ts2R9//IGYmBi0b98eMpkMZ86cgVqthpeXl9SlYfHixaioqIBcLoejoyMOHz6MGzduGOXaLVq0wMGDB7Fx40bMnDmT4UdVGHxUKzg7OyMpKQlXrlzB6NGjq5bG6NEyMzMxbtw4dO7cGe7u7jh//jw+/fRTNGrUSOrSqqxfvx5XrlxB06ZN8cUXX2DTpk2oV6+e0a7fqFEjpKen48cff8SECRP4c0MAuLmFapmSkhIMGjQIzs7O2LBhA0dUPcTp06fxwQcfIDU1FdOmTcO0adPg7u4udVmP5eHhgbNnz6JBgwYmuX5RUREGDRoER0dHfPvtt7C3tzfJfah2YMdHtYqDgwO2b9+O0tJSDBkyBGVlZVKXZDGOHz+OQYMGITg4GD4+Prh8+TLeeustiw89jUaD3Nxco3Z6D3JycsKOHTugVCrRp08fFBQUmOxeZPkYfFTrqFQqbN68GQqFAgMHDkRJSYnUJUnq0KFDiIiIQGRkJPz9/XH58mW88cYbcHZ2lro0vfz9999wc3ODXG7aE9RVKhW+/fZbtGzZEgEBAbh9+7ZJ70eWi8FHtZKdnR2+++47uLm5oX///iguLpa6JLMSRRF79+5FQEAAoqKiEBkZiUuXLuGVV16Bo6Oj1OUZJDs722RLnA+qPDopODgYfn5+uHbtmlnuS5aFwUe1lkKhwLp169C4cWNERESgsLBQ6pJMThRFJCUloUePHoiJicGYMWOQmZmJSZMmQaWqnefSZWdno379+ma7nyAI+OCDDzBhwgT4+fkhMzPTbPcmy8Dgo1pNLpdj9erV8Pb2RmhoKPLy8qQuySR0Oh22bduG5557Dm+88QZefvllnD17FmPGjKn1G3zMHXyVZsyYgbfeegu9evXCiRMnzH5/kg6Dj2o9mUyGZcuWoVOnTggODkZOTo7UJRmNVqvFt99+iw4dOuD999/HvHnzcOrUKQwbNszkz8TMJSsrS5LgA4Dx48cjLi4OYWFhSE9Pl6QGMj8GH1kFmUyGxYsXo2fPnlaxcaGiogJr1qxBmzZtEBcXh48//hjHjh3DgAEDDDoLrzaQquOrNGjQIHz77bcYPHgwEhISJKuDzMe6/gaRTRMEAYsWLUJERAR69+6NW7duSV2SwUpLS7Fs2TK0atUKGzZswIoVK3Dw4EGEhYVZ7ZBuc25ueZTAwEDs3LkTL730EtavXy9pLWR6PJaIrIogCFiwYAFUKhV69eqFtLS0J04q0YoijudokZSlwS95WhRoRGhFQP7P8TjPusoR3kCBzu6PPh6npoqLi7FixQosWrQIPj4++Pbbb01yFp4lkrrjq9S1a1fs3bsXYWFhuHPnDl5++WWpSyITYfCR1REEAW+99Rbs7Ozg7++PvXv3omnTpv/zdVllOnxwz4GoJY+Yf739Lw0+ulCGFnVkCPS4eyBqA5VxFkvy8/OxdOlSfPbZZ+jevTsSEhLQqVMno1y7trCU4AOAtm3bIiMjA8HBwfj7778xf/58q+20bRmDj6zW7NmzoVKp4O/vj7S0NDRr1gzA3Q7v7d9Lse56Ba6W6Dexr1gHnM7X4XR+ObberEB0EyXebWMPWTV/Kebk5GDx4sWIi4tDSEgI0tLSJDsWSGqWFHwA4OXlhYyMjKrO74svvrC656q2jv9vklWbPn06ZsyYgV69euHixYu4XKRFj4xCfHSxXO/Qe9DVEhEfXSxH94xCXC4y7KSIrKwszJkzBy1btsTVq1dx6NAhbNiwwWZDD5B2V+ejeHp6Yv/+/Th16hSio6NRUVEhdUlkRBxSTTZh5cqV+L8VG+H6/jZcMOKQF28nGTZ1cUAH18cvnty4cQOLFi3CN998g+HDh2PWrFkWcSyQ1LRaLVQqFUpLS6FQWN4CVElJCYYOHQpRFBEfH1/rpuLQw7HjI5sQMGIcHOdvMmroAUBmkQ5DjpU8svP7448/MGXKFDzzzDMQBAFnzpzBkiVLGHr/uHPnDlxdXS0y9IC7Q9G3bt0KNzc3hIaGIjc3V+qSyAgYfGT1tKKIqOPF+KPCNL9cM4t0GHW8BNp7Fk8uXLiA8ePHo3PnznBzc8O5c+cs7iw8S2Bpz/ceRqlUYu3atfDx8am1r8nQ/Rh8ZPXe/r0Ux3MfsWXTSH7O1WL+uTKcOXMGI0eOxAsvvAAvLy9cvHgRH3zwgeTvqVmq2hB8wN0BCV988QUGDBgAPz8//PHHH1KXRDXA4COrllWmw7rrFagw8ZPsChFYcvZvBAwYgo4dO+LSpUt4++23Lf4sPKllZWXVmg8FgiDg7bffxrRp0+Dn54ezZ89KXRJVk2UurBMZycLMsmrv3jTUHUUdTN10FLN8asc5eJagtnR896o80T4gIAA7duxA165dpS6JDMSOj6yWVhSRmq3R++ud5MB/gp0xovF/TzuoowCuBDvjxYb6fUZMz8F9z/ro8Wpj8AHAqFGjsHLlSvTt2xepqalSl0MGYvCR1Tqeo8WlIv2f7RVpgX//UoIv2tvDw+7ui+kftbXHz7labLmpX4BeKtLhRK5h7/bZstoafADQt29fbN68GSNHjsSWLVukLocMwOAjq5WUpXnkGLJHScnWYOctDRa3t4d/PTmGNlYi9tcSvb+/WAfsvqV/l2nranPwAUDPnj2xZ88eTJs2DatWrZK6HNITn/GR1TqVX73Oa/qZEpwNcEZwfQVm/FaKv8oMW7o8mceOT1+1aXPLo/j4+GD//v0ICQnBnTt3MHPmTKlLoidgx0dWK7+aWzlzK4DfCnRwlAvYesPwUVUFGj7j01dt7/gqeXt74+DBg1izZg1mz54NDsSybAw+slraav7uiWqixP9zEJCarcGHbe0N/n7mnv6sJfgAoEmTJjhw4AD27t2LSZMmQatl52+pGHxkteTVODihvp2Az56xx0unSjDpVAmGNlbCr57coGsoeIqNXnQ6He7cuQMPDw+pSzEaDw8PpKWl4dKlSxgxYgTKysqkLokegsFHVstFaXgCxXWwx/abGuy/rcVfZSLe+K0UXz3rADsD/qY4M/n0kpOTgzp16kCpVD75i2sRZ2dn7Ny5ExqNBv369UNhYaHUJdEDGHxktTq6GNapRf5LgR71FJj52393ca66WoHrJSLeelql93V8XA27r62yho0tj2Jvb4/4+Hg0bdoUQUFBuHPnjtQl0T0YfGS1whso4GDAT/gPf2nQeE8B8h54GyHoUBHm/q7fkpWjDAjz5GZpfVjT872HUSgUWLlyJfz8/NCzZ0/cuHFD6pLoHww+slqd3eVo4WTeH/EWTjJ0cmPHpw9rDz7g7nzPjz76CKNGjUKPHj1w8eJFqUsi8D0+smJyQUBQfQXOFJSb7Z6B9RWQC3zGpw9bCD7gbvjNnj0bdevWhb+/P3bt2oWOHTtKXZZNY8dHVm2OtwpPOZgniLwcBLzprf+zQFtnK8FX6d///jc+++wzhISE4Mcff5S6HJvG4COr1kAlQ3QTJaqxwdMgSgEY3dQO9VX8K6Wv7Oxsq93c8ihDhw7F2rVrMWDAACQlJUldjs3i31Kyeu+0sUdnN9P+qD/nJsf81uz2DJGVlWVTHV+l0NBQ7NixA2PHjsW3334rdTk2icFHVk8uCNjQ2RHeJtro4u0kw4bODpDx2Z5BbG2p816+vr5IS0vDzJkzsWTJEqnLsTkMPrIJzZ3k2NTFwejh18pJQPukRUDWVaNe1xbYcvABwDPPPIMDBw7gk08+wYIFCzjf04wYfGQzOrgqkOTrCF93eY2f+SkFwNddjh2dFNj62QK0bNkSnTp1wurVq1FQUGCcgq2cLT7je1Dz5s1x8OBBxMfH47XXXoNOZ+A5WlQtgsiPGWRjtKKI+efKsO5aOa6UGP7j7+UgILqpHd5prYJMEODv748DBw4AuDuxQ6VS4fbt21Ao+LbQo+h0OqhUKhQVFcHOzk7qciSXk5ODvn37omXLlli1ahV/dkyMHR/ZHLkg4L029jjqXwevNrdDe2cZHJ/wN8FRBrR3lmF6czsc86+D99rYVz3Ti46OhqOjIwBAq9Vi6dKl/MX1BLm5uXBycmLo/cPd3R3JycnIysrCiy++iNLSUqlLsmrs+MjmaUURJ3K12H1Lg5N5WhRoRGjEu6csOCsE+LjKEeapQCc3+UNfTr958ya8vLygUCjg6OiIzz//HKNGjZLgT1J7nD9/Hn379sWFCxekLsWilJeXY8yYMbh58yZ27NgBFxcXqUuySvxYSjZPLgjo4q5AF/fq/XVo2LAhxo4di1GjRsHDwwMBAQFwcXFB//79jVyp9bD1jS2PYmdnh/Xr12PatGno3bs3du/ezf9OJsClTiIjWLFiBXr27Im2bdsiMTEREydOxL59+6Quy2JxY8ujyeVyqNVqREREwM/PD1evcsewsTH4iIzsueeew6ZNmzBs2DAcPXpU6nIsEju+xxMEAe+99x4mTZoEPz8/nDt3TuqSrAqDj8gE/P39sXr1avTv3x9nzpyRuhyLY6tTWww1ffp0vPvuu+jduzd+/vlnqcuxGgw+IhPp27cvPvvsM4SFheHy5ctSl2NR2PHpb8yYMVi2bBkiIiK4fG4kDD4iExoxYgTmzp2L4OBgHkR6DwafYSIjI/H9999j2LBh2L59u9Tl1Hrc1UlkYpMnT0Zubi6Cg4Nx4MAB1KtXT+qSJMfgM1zv3r2RlJSEvn37Ijc3F2PHjpW6pFqLwUdkBrNnz0Zubi7Cw8ORlpYGZ2dnqUuSFHd1Vk/nzp2xb98+hIaGIicnB9OnT5e6pFqJS51EZrJw4UJ06tQJ/fv3R0lJidTlSIqbW6qvdevWyMjIwPLlyzF37lwOt64GTm4hMiOtVotRo0ahqKgIW7ZsgVKplLoksxNFESqVCgUFBVCpeIZhdWVnZyMsLAzPP/88vvzyS8jlcqlLqjXY8RGZkVwux9q1a6HT6TBu3DibnMafl5dXNcybqq9+/frYt28fzp49i6ioKJSXl0tdUq3B4CMyM6VSiU2bNuHatWuYNm2azS1V8fme8bi4uGD37t0oKSlBZGQkiouLpS6pVmDwEUnAwcEBCQkJOHLkCObNmyd1OWbFHZ3GZW9vjy1btsDT0xPBwcHIycmRuiSLx+Ajkkjlp/UtW7Zg0aJFUpdjNtzYYnwKhQKrV69G165d4e/vj5s3b0pdkkVj8BFJyMPDAykpKViyZAm++uorqcsxC3Z8piGTyfDpp59i6NCh8PPz47Sgx+B7fEQSa9KkCVJSUuDv7w8XFxcMGzZM6pJMisFnOoIgYO7cuahbty569uyJpKQktG/fXuqyLA6Dj8gCtGzZEklJSQgODoazszMiIiKkLslksrOz0bRpU6nLsGpTpkyBu7s7goKCsH37dvj6+kpdkkXhUieRhejQoQN++OEHjB07FhkZGVKXYzLs+MxjxIgRWLNmDfr374/k5GSpy7EoDD4iC9KtWzds3LgRL774Ik6cOCF1OSbBzS3mExERgW3btiE6Ohrx8fFSl2MxGHxEFiYoKAjLly9Hnz59rPIAUnZ85tWjRw8kJydj+vTpWLFihdTlWAQ+4yOyQAMHDkR+fj5CQkKQkZEBLy8vqUsyGgaf+XXs2BHp6ekICQnBnTt3MGvWLAiCIHVZkmHwEVmoMWPGIC8vD8HBwcjIyICnp6fUJdWYKIoMPom0bNkSBw8eREhICP7++2989NFHNht+XOoksmAvv/wyRo0ahZCQEKuYyFFQUAA7Ozs4ODhIXYpNatSoEQ4cOICDBw9i4sSJ0Gg0UpckCQYfkYWbN28eAgMD0adPHxQVFUldTo2w25Ne3bp1kZKSgmvXrmHo0KEoLS2VuiSzY/ARWThBEPDJJ5+gdevWGDhwIMrKyqQuqdq4o9My1KlTBwkJCVAoFOjTpw8KCgqkLsmseB4fUS2h0WgwfPhwiKKI77//HgpF7XtEv2PHDqxYsQKJiYlSl0K4ez5kTEwMfvnlF+zatQseHh5P/h5RxPEcLZKyNPglT4sCjQitCMgFwFkh4FlXOcIbKNDZXQ65hT5DZPAR1SJlZWXo378/GjVqhFWrVkEmq12LNqtWrcKPP/6I1atXS10K/UMURbz55pv44YcfkJycjCZNmjz067LKdPggswxp2RpcKtKh5DFHSTrKgBZ1ZAj0UGCOtwoNVJb1c2pZ1RDRY6lUKmzduhWZmZl47bXXat1ZfnzGZ3kEQcDChQsxbtw49OjRA5mZmff9e60oYu7ZEnRJL8QXl8txpuDxoQcAxTrgdL4On18uR5f0Qsw9WwKdBf2sMviIahknJyfs3LkT+/fvx7vvvit1OQZh8FmumTNnYt68efD396+aGnS5SIseGYX46GI5rpZUL7iuloj46GI5umcU4nKR1pglV1vte0hARHBzc8OePXvg5+cHNzc3vPLKK1KXpJesrCx06NBB6jLoESZMmAB3d3eEhYVh89GzeOmiCplFT2jv9FAhAj/l6BB+uBibujigg6u00cPgI6qlPD09kZKSAj8/P7i6umLs2LFSl/RE7Pgs36BBg+DZ4XmMv2hnlNC7V2aRDkOOlSDJ1xHNneRGvbYhuNRJVIt5eXkhOTkZc+bMwdatW6Uu54kYfJZPK4qYccMFmUWmeSaXWaTDqOMl0Er4zI/BR1TLtW7dGrt27cLkyZORkpIidTmPlZ2djQYNGkhdBj3G27+X4niucTu9B/2cq8X8c9K9j8rgI7ICPj4+2LJlC0aOHInDhw9LXc5DcU6n5csq02Hd9QpUmLgZqxCBddfKkV1m2oB9FAYfkZXw8/PD2rVrMWDAAPz6669Sl/M/CgsLIZPJ4OjoKHUp9AgLM8uqvXvTUFdKRCzMlKbrY/ARWZHw8HB8+eWXCA8Px4ULF6Qu5z7s9iybVhSRmm3Y0OrudeX40c8JuREu+DvcGQd7OOE5N/03raRmayR51sddnURWZujQocjPz686zqhp06ZSlwSAwWfpjudoccmAXZzOCiCxmxNiTpUg/s8K2MkAv3oKlOn0D7JLRTqcyNWii7t5o4jBR2SFJk6ciNzc3Krws4TA4cYWy5aUpXniRJZ7ede529l992cFAKBUB6QY2DEW64DdtzRmDz4udRJZqRkzZmDw4MEICwtDXl6e1OWw47Nwp/INm6qSWaiFVhTxtY8Dwhoo4Kas3n1P5pl/mguDj8iKvffee/D19UW/fv1QXFwsaS08ksiy5Ru4lbNAA/TIKIII4KtnHZAd5oIfujqigcqwExkKNOZ/xsfgI7JigiBg8eLF8PLywuDBg1FeXi5ZLez4LJu2GvlzrlCHcSdL0DS5AM/sK0QjBwGfP2Nv0DUkyD0GH5G1k8lkWL16NZRKJUaPHg2tVppBwQw+yyav4dF55wt1+PpqBZ5xMWwUmUKCI/sYfEQ2QKlU4vvvv0dWVhZiYmIkOc6Im1ssm4vSsAR6uo4Mr7WwQ2P7u9/XxF7AiMZK/HTHsA9WzhIkH4OPyEbY29vjhx9+wKlTpzBr1iyzhx87PssjiiLOnTuHxYsXQ7hy1qDvLdCIeN5djiM966Cwjwt+6lkHZwq0eP23EoOu4+Nq/mHVfJ2ByIY4Oztj165d8Pf3h7u7O+bMmWO2e3Nzi2X4+++/kZaWhuTkZCQnJwMAQkJCEN6hB5IKoPcrDTdKRQz72bCQe5CjDAjzNH8MCWJtO8KZiGrsxo0b8PPzw4wZMxATE2OWezo6OiIrKwt16tQxy/3orvLychw+fLgq6M6fPw9/f3+EhIQgJCQE3t7eEAQBWlHEs/sKcabAfPMz2zvLcLJ3HcgF8y53suMjskGNGjVCamoqevbsCRcXF0RFRZn0fkVFRRBFEU5OTia9D91dvjx//jxSUlKQnJyMAwcOwNvbGyEhIfj444/h6+sLOzu7//k+uSAgqL4CZwrMt/M3sL7C7KEHMPiIbFazZs2we/duBAYGwsXFBf369TPZvSo3tggS/JKzBQ8uX4qiiNDQUERHR+Prr79GvXr19LrOHG8Vtt6sMMugai8HAW96q0x+n4dh8BHZsHbt2iEhIQERERGIj49H7969TXIfbmwxrsrly8qu7vz58/Dz80NISAhmzJiBp59+ulofMhqoZIhuosRHF8tNejSRUgBGN7VDfZU0+ysZfEQ2rkuXLoiPj8ewYcOwc+dOdOnSxej34MaWmhFFEZmZmVUd3b3Ll4sWLXrk8mV1vNPGHmm3Nfgpx3TP+p5zk2N+a2m6PYDBR0QAevfujVWrVqFfv35IS0tDu3btjHp9dnyGu3Pnzn3LlzqdDqGhoRg1ahTWrFkDDw8Pk9xXLgjY0NkR4YeLkWnAaQ368naSYUNnB8gkXPZm8BERAKBfv3745JNPEBoaigMHDqB58+ZGuzaD78nKy8vx008/VQXduXPn0LNnT4SEhOD111+v9vJldTR3kmNTFwcMOVZi1PDzdpJhUxcHNHMy/7t792LwEVGVqKgo5OXlVR1n1KhRI6Ncl1Nb/lfl8mXlc7r09HR4e3sjODgYH330EXx9faFSSbcc2MFVgSRfR4w6XoKfc7U1euanFO4ub27oLH3oAQw+InrAlClTkJeXh5CQEKSnp+u9I/BxsrOz0bp1ayNUV7s9bPkyJCQEUVFRWL16tcmWL6uruZMcGX5OmH+uDOuuleNKNXZ7ejkIiG5qh3daqyRd3rwXg4+I/sfs2bORk5ODiIgIpKamwtnZuUbXs9XNLfcuX6akpOD333+v2n352muvoXXr1hb/iodcEPBeG3tMa26HhZllSMvW4FKRDsWPWQF1lAEtnGQIqq/AHG+VZLs3H4WTW4jooURRxKRJk3Dx4kXs2rUL9vaGHTdzr65du2Lx4sXo1q2bESu0PKIo4sKFC1UdXXp6Olq1alU1JUXq5Utj0IoiTuRqsfuWBifztCjQiNCId09ZcFYI8HGVI8xTgU5uckleTtcHg4+IHkmr1SIqKgolJSXYvHkzlMrqHbPdrFkzpKamokWLFkauUHqVy5eVz+o0Gk1V0AUGBtpkp2vpGHxE9Fjl5eUYMGAA6tWrh2+++QYymeHLVq6urrh27RpcXFxMUKF5VVRU3Lf78t7ly5CQkFqxfGnrGHxE9ETFxcUICwtDhw4d8OWXXz7yF7tWFHE8R4ukLA1++WcZTCsCRQUFaFzXBc+6yhHeQIHO7pa7DPYgURRx8eLFqqDbv38/WrZsWRV0L7zwQq1fvrQ1DD4i0kteXh4CAgIQHh6OBQsW3Pfvssp0+OCejQ+PO9rGUQa0qCNDoMfdjQ8NLGzjAwDk5OTct/uycvkyODgYQUFBXL6s5Rh8RKS37Oxs9OzZExMmTMCMGTOgFUW8/Xsp1l2v3mDjpxwERDdR4t029pJuda+oqMCRI0eqgu7s2bPo0aNHVVfXpk0bLl9aEQYfERnk+vXr8PPzw+zPl+Fr9xdwPFdX45ebO7vJsKGzI5qb6eXmB5cv09PT0aJFi6qurnv37ly+tGIMPiIyWNqFm/j3JRUulxkvqCrHWXVwNc3rxTk5Odi7d29V2JWXl9+3+5KTZWwHg4+IDHK5SGvSAcZJvsbp/CqXLytfM/jtt9/Qo0cPBAcHIyQkBG3btuXypY1i8BGR3rSiiB4ZhSY9ssbX/e6YLEN3fYqiiEuXLt23+7J58+b37b6syUv4ZD0YfESkt7lnS8xySOmsViq81+a/ISWKIt5++21kZWVh2bJlVf88Nzf3vuXLsrKy+3ZfcvmSHobBR0R6ySrToUt6YbV2bxrKy0HAMf86qK+SQaPRYPz48di8eTMEQUBSUlLVqwZnzpy5b/clly9JHww+ItLL9NMl+Pxyufnu19wOC1qICAoKwvHjx1Fefvfe3t7eGDBgAEJCQtC9e3cuX5LBeDoDET2RVhSRmq0x6HvGNFXi9ZYqtHCUIV8jYuvNCsw5W4p8PS+Tmq3B0di+OHz4MOTyu5td5HI5oqKi8NZbbxn6RyCqwuAjoic6nqPFJQN2cb7Wwg5vtFRhzMkSpGVr0NhewJKODkh+wQk9Moqg0WOd6VKRDnvid8Hu6m84ceIEMjIy8NNPP+HGjRs1+JMQcamTiPTwzrlSzD9fptfXOiuAG6EuGH+yBJtuVFT9cyc5cDnYGTN/K8XaaxWPucJ/vfu0CvNacymTjMvyhuQRkcU5la/V+2tfqKuAvQzYevP+cCvSAkm3NAipr/9C08k8/e9LpC8GHxE9Ub4B7y942Am4XX73VIYH3SzTob5K/12XBfqsiRIZiMFHRE/0sBB7lNvlIjzsBMgfkm8NVTJkl+l/MeYemQKDj4ie6GEh9iiH72hQpgMGNbz/tHZHORDuqUD63/rvDlXwlTwyAQYfET2Ri1L/BMrXAO+cL8WX7e0R2kABhXD3hfRNXRxxu1zEhuv6bWwBAGcmH5kAX2cgoifq6CLHtpv6d2qLLpbj73IRH7ezR0snGezlAvbf1iDoUBGKDdiv4uNqnmOKyLaw4yOiJwpvoICDgb8tVl+tQPt9hXBIzMf4k8Vo4SQzaOnSUQaEefKzORkfg4+InqizuxwtnKr/62LN1Qq8ebYUL9TVP8haOMnQyY0dHxkfP04R0RPJBQFB9RU4U1D9WZ3rDXi2BwCB9RUGH01EpA9ObiEivUh1OgORsfGnioj00kAlQ3QTJQzY4FktSgEY3dSOoUcmw46PiPRmrhPYD/o5QcZlTjIRfqQiIr3JBQEbOjvCuwYbXR7H20mGDZ0dGHpkUgw+IjJIcyc5NnVxMHr4eTvJsKmLA5o5cScnmRaXOomoWi4XaTHqeAl+ztXCgBnW/0MpAM+5ybGhM0OPzIPBR0TVphVFzD9XhnXXynGlGrs9vRwERDe1wzutVVzeJLNh8BFRjWWV6bAwswxp2RpcKtKh+DF7Xxxld19OD6qvwBxvFXdvktkx+IjIaLSiiBO5Wuy+pcHJPC0KNCI04t1TFpwVAnxc5QjzVKCTm5wvp5NkGHxERGRTuMZAREQ2hcFHREQ2hcFHREQ2hcFHREQ2hcFHREQ2hcFHREQ2hcFHREQ2hcFHREQ2hcFHREQ2hcFHREQ2hcFHREQ2hcFHREQ2hcFHREQ2hcFHREQ2hcFHREQ25f8DEy6Fxx1/Oh0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "nx.draw(\n",
    "    graph_nx, \n",
    "    with_labels=True, \n",
    "    node_size=900,\n",
    "    font_color='white', \n",
    "    node_color=COLORS[0]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 478,
   "metadata": {},
   "outputs": [],
   "source": [
    "causal_model = gcm.InvertibleStructuralCausalModel(graph_nx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 479,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Fitting causal mechanism of node P: 100%|███████████████████████████████████████████████| 5/5 [00:00<00:00, 416.72it/s]\n"
     ]
    }
   ],
   "source": [
    "causal_model.set_causal_mechanism('S', gcm.EmpiricalDistribution())\n",
    "causal_model.set_causal_mechanism('X', gcm.AdditiveNoiseModel(gcm.ml.create_linear_regressor()))\n",
    "causal_model.set_causal_mechanism('Y', gcm.AdditiveNoiseModel(gcm.ml.create_linear_regressor()))\n",
    "causal_model.set_causal_mechanism('P', gcm.AdditiveNoiseModel(gcm.ml.create_linear_regressor()))\n",
    "causal_model.set_causal_mechanism('Q', gcm.AdditiveNoiseModel(gcm.ml.create_linear_regressor()))\n",
    "\n",
    "gcm.fit(causal_model, df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 480,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{('Q', 'Y'): 0.0006632517083816319,\n",
       " ('S', 'Y'): 0.008486091866545382,\n",
       " ('X', 'Y'): 0.006866684034567223}"
      ]
     },
     "execution_count": 480,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gcm.arrow_strength(causal_model, 'Y')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 481,
   "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>S</th>\n",
       "      <th>Q</th>\n",
       "      <th>X</th>\n",
       "      <th>Y</th>\n",
       "      <th>P</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.4</td>\n",
       "      <td>0.21</td>\n",
       "      <td>0.539558</td>\n",
       "      <td>0.169005</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     S    Q     X         Y         P\n",
       "0  0.5  0.4  0.21  0.539558  0.169005"
      ]
     },
     "execution_count": 481,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gcm.counterfactual_samples(\n",
    "    causal_model,\n",
    "    {'X': lambda x: .21},\n",
    "    observed_data=pd.DataFrame(data=dict(X=[.5], Y=[.75], S=[.5], Q=[.4], P=[.34])))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:causal_book_py38]",
   "language": "python",
   "name": "conda-env-causal_book_py38-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}