examples/statistical_rethinking_lectures/06-Good_&_Bad_Controls.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "98327666-ff00-4e30-95cd-97364f288b3f",
   "metadata": {},
   "source": [
    "(lecture_06)=\n",
    "# Good & Bad Controls\n",
    "\n",
    ":::{post} Jan 7, 2024\n",
    ":tags: statistical rethinking, bayesian inference, causal inference, controls\n",
    ":category: intermediate\n",
    ":author: Dustin Stansbury\n",
    ":::\n",
    "\n",
    "This notebook is part of the PyMC port of the [Statistical Rethinking 2023](https://github.com/rmcelreath/stat_rethinking_2023) lecture series by Richard McElreath.\n",
    "\n",
    "[Video - Lecture 06 - Good & Bad Controls](https://youtu.be/NSuTaeW6Orc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c1ffbccd",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Ignore warnings\n",
    "import warnings\n",
    "\n",
    "import arviz as az\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import pymc as pm\n",
    "import statsmodels.formula.api as smf\n",
    "import utils as utils\n",
    "import xarray as xr\n",
    "\n",
    "from matplotlib import pyplot as plt\n",
    "from matplotlib import style\n",
    "from scipy import stats as stats\n",
    "\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "# Set matplotlib style\n",
    "STYLE = \"statistical-rethinking-2023.mplstyle\"\n",
    "style.use(STYLE)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "128e5650",
   "metadata": {},
   "source": [
    "# Avoid being clever at all costs\n",
    "Being clever is\n",
    "- unreliable\n",
    "- opaque\n",
    "\n",
    "Using explicit causal models allows one to:\n",
    "- derive implications using logic\n",
    "- verify work & assumptions\n",
    "- facilitates peer review & verification"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a99c2b9b",
   "metadata": {},
   "source": [
    "# Confounds\n",
    "## Review\n",
    "- The Fork $X \\leftarrow Z \\rightarrow Y$\n",
    "  - $X$ adn $Y$ are associated unless we stratify by $Z$\n",
    "- The Pipe $X \\rightarrow Z \\rightarrow Y$\n",
    "  - $X$ adn $Y$ are associated unless we stratify by $Z$\n",
    "- The Fork $X \\rightarrow Z \\leftarrow Y$\n",
    "  - $X$ adn $Y$ are not associated unless we stratify by $Z$\n",
    "- The Descendant $Z \\rightarrow A$\n",
    "  - Descendent $A$ takes on behavior of parent $Z$\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "310fca8a",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"494pt\" height=\"104pt\"\n",
       " viewBox=\"0.00 0.00 493.66 103.89\" 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 99.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-99.89 489.66,-99.89 489.66,4 -4,4\"/>\n",
       "<!-- U -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>U</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"71.24\" cy=\"-19.45\" rx=\"68.59\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"71.24\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">confound, U</text>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"249.72\" cy=\"-43.45\" rx=\"71.24\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"249.72\" y=\"-39.9\" font-family=\"Times,serif\" font-size=\"14.00\">treatment, X</text>\n",
       "</g>\n",
       "<!-- U&#45;&gt;X -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>U&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M133.47,-27.77C146.54,-29.55 160.48,-31.44 173.99,-33.28\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"173.32,-36.72 183.7,-34.6 174.27,-29.79 173.32,-36.72\"/>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"421.31\" cy=\"-19.45\" rx=\"64.35\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"421.31\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">outcome, Y</text>\n",
       "</g>\n",
       "<!-- U&#45;&gt;Y -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>U&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M139.39,-16.53C152.36,-16.08 165.84,-15.68 178.48,-15.45 241.8,-14.25 257.65,-14.17 320.96,-15.45 329.03,-15.61 337.47,-15.84 345.86,-16.12\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"345.63,-19.62 355.74,-16.47 345.87,-12.62 345.63,-19.62\"/>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M313.39,-34.58C325.67,-32.84 338.62,-31.01 351.1,-29.24\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"351.49,-32.72 360.9,-27.85 350.5,-25.79 351.49,-32.72\"/>\n",
       "</g>\n",
       "<!-- unmeasured -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>unmeasured</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"71.24\" cy=\"-76.45\" rx=\"71.24\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"71.24\" y=\"-72.9\" font-family=\"Times,serif\" font-size=\"14.00\">unmeasured</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5b0e145520>"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[(\"U\", \"X\"), (\"U\", \"Y\"), (\"X\", \"Y\")],\n",
    "    node_props={\n",
    "        \"X\": {\"label\": \"treatment, X\"},\n",
    "        \"Y\": {\"label\": \"outcome, Y\"},\n",
    "        \"U\": {\"label\": \"confound, U\", \"style\": \"dashed\"},\n",
    "        \"unmeasured\": {\"style\": \"dashed\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "78c4b9b2",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"539pt\" height=\"161pt\"\n",
       " viewBox=\"0.00 0.00 539.27 160.89\" 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(4 156.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-156.89 535.27,-156.89 535.27,4 -4,4\"/>\n",
       "<!-- U -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>U</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"94.05\" cy=\"-19.45\" rx=\"68.59\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"94.05\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">confound, U</text>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"295.33\" cy=\"-66.45\" rx=\"71.24\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"295.33\" y=\"-62.9\" font-family=\"Times,serif\" font-size=\"14.00\">treatment, X</text>\n",
       "</g>\n",
       "<!-- U&#45;&gt;X -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>U&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"lightgray\" d=\"M147.93,-31.92C172.95,-37.82 203.06,-44.92 229.45,-51.14\"/>\n",
       "<polygon fill=\"lightgray\" stroke=\"lightgray\" points=\"228.5,-54.52 239.04,-53.4 230.11,-47.7 228.5,-54.52\"/>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"466.92\" cy=\"-42.45\" rx=\"64.35\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"466.92\" y=\"-38.9\" font-family=\"Times,serif\" font-size=\"14.00\">outcome, Y</text>\n",
       "</g>\n",
       "<!-- U&#45;&gt;Y -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>U&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M161.35,-23.56C226.25,-27.58 324.9,-33.7 392.23,-37.88\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"391.88,-41.36 402.07,-38.49 392.31,-34.37 391.88,-41.36\"/>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M359,-57.58C371.28,-55.84 384.23,-54.01 396.71,-52.24\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"397.09,-55.72 406.5,-50.85 396.11,-48.79 397.09,-55.72\"/>\n",
       "</g>\n",
       "<!-- R -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>R</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"94.05\" cy=\"-76.45\" rx=\"94.05\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"94.05\" y=\"-72.9\" font-family=\"Times,serif\" font-size=\"14.00\">randomization, R</text>\n",
       "</g>\n",
       "<!-- R&#45;&gt;X -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>R&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M185.94,-71.88C195.11,-71.42 204.37,-70.96 213.43,-70.51\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"213.43,-74.01 223.24,-70.01 213.08,-67.02 213.43,-74.01\"/>\n",
       "</g>\n",
       "<!-- unmeasured -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>unmeasured</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"94.05\" cy=\"-133.45\" rx=\"71.24\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"94.05\" y=\"-129.9\" font-family=\"Times,serif\" font-size=\"14.00\">unmeasured</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5b0e670380>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[(\"U\", \"X\"), (\"U\", \"Y\"), (\"X\", \"Y\"), (\"R\", \"X\")],\n",
    "    node_props={\n",
    "        \"X\": {\"label\": \"treatment, X\"},\n",
    "        \"Y\": {\"label\": \"outcome, Y\"},\n",
    "        \"U\": {\"label\": \"confound, U\", \"style\": \"dashed\"},\n",
    "        \"R\": {\"label\": \"randomization, R\"},\n",
    "        \"unmeasured\": {\"style\": \"dashed\"},\n",
    "    },\n",
    "    edge_props={(\"U\", \"X\"): {\"color\": \"lightgray\"}},\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb56f557",
   "metadata": {},
   "source": [
    "The gold standard is **Randomization**. However randomization often this generally isn't possible:\n",
    "- impossible\n",
    "- pragmatism\n",
    "- ethical concerns\n",
    "- unmeasured confounds\n",
    "\n",
    "## Causal Thinking\n",
    "- We would like a procedure $do(X)$ that intervenes on $X$ in such a way that it can \"mimic\" the effect of randomization.\n",
    "- Such a procedure would transform the Confounded graph:\n",
    "\n",
    "#### Without randomization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "6a807a32",
   "metadata": {
    "jupyter": {
     "source_hidden": true
    }
   },
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"242pt\" height=\"71pt\"\n",
       " viewBox=\"0.00 0.00 242.00 70.89\" 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 66.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-66.89 238,-66.89 238,4 -4,4\"/>\n",
       "<!-- U -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>U</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">U</text>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"117\" cy=\"-43.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"117\" y=\"-39.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- U&#45;&gt;X -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>U&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M52.51,-26.13C61.12,-28.48 70.97,-31.16 80.3,-33.71\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"79.28,-37.06 89.85,-36.31 81.13,-30.31 79.28,-37.06\"/>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"207\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"207\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- U&#45;&gt;Y -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>U&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M54.23,-17.38C65.26,-16.61 78.25,-15.82 90,-15.45 113.99,-14.67 120.01,-14.67 144,-15.45 151.9,-15.7 160.35,-16.14 168.39,-16.63\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"168.05,-20.11 178.26,-17.28 168.51,-13.13 168.05,-20.11\"/>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M142.51,-36.76C151.12,-34.41 160.97,-31.73 170.3,-29.18\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"171.13,-32.58 179.85,-26.58 169.28,-25.83 171.13,-32.58\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bd01466c0>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    [\n",
    "        (\"U\", \"X\"),\n",
    "        (\"U\", \"Y\"),\n",
    "        (\"X\", \"Y\"),\n",
    "    ],\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a2b28cc6",
   "metadata": {},
   "source": [
    "in such a way that all the non-causal arrows entering X have been removed"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6a210218-cdda-48ea-80ff-300f3ec1b39a",
   "metadata": {},
   "source": [
    "#### With \"randomization\" induced by $do(X)$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "bb74840b",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"198pt\" height=\"104pt\"\n",
       " viewBox=\"0.00 0.00 197.75 103.89\" 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 99.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-99.89 193.75,-99.89 193.75,4 -4,4\"/>\n",
       "<!-- U -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>U</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-76.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-72.9\" font-family=\"Times,serif\" font-size=\"14.00\">U</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"162.75\" cy=\"-47.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"162.75\" y=\"-43.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- U&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>U&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M53.73,-73.14C71.78,-70.58 96.4,-66.6 117.75,-61.45 120.89,-60.69 124.13,-59.82 127.37,-58.9\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"128.17,-62.31 136.73,-56.07 126.14,-55.61 128.17,-62.31\"/>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M53.47,-24.78C73.83,-29.04 102.63,-35.07 125.41,-39.84\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"124.39,-43.2 134.89,-41.82 125.82,-36.35 124.39,-43.2\"/>\n",
       "<text text-anchor=\"middle\" x=\"94.88\" y=\"-43.65\" font-family=\"Times,serif\" font-size=\"14.00\">do(X) &#160;&#160;</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5b0e1644a0>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    [\n",
    "        (\"U\", \"Y\"),\n",
    "        (\"X\", \"Y\"),\n",
    "    ],\n",
    "    edge_props={(\"X\", \"Y\"): {\"label\": \"do(X)   \"}},\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb7fc683-405d-4af1-ab8b-9899b0ee46cf",
   "metadata": {},
   "source": [
    "It turns out that we can analyze graph structure to determine if there is such a procedure that exists"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "098982b1",
   "metadata": {},
   "source": [
    "## Example: Simple Confound"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "3cb4d99a",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"242pt\" height=\"70pt\"\n",
       " viewBox=\"0.00 0.00 242.00 69.89\" 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 65.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-65.89 238,-65.89 238,4 -4,4\"/>\n",
       "<!-- U -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>U</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"27\" cy=\"-42.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-38.9\" font-family=\"Times,serif\" font-size=\"14.00\">U</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"207\" cy=\"-42.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"207\" y=\"-38.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- U&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>U&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M54.22,-45.03C65.25,-46 78.24,-46.97 90,-47.45 113.98,-48.41 120.02,-48.41 144,-47.45 151.9,-47.13 160.36,-46.58 168.4,-45.97\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"168.59,-49.46 178.27,-45.15 168.02,-42.49 168.59,-49.46\"/>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"117\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"117\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- U&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>U&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M52.98,-35.92C61.51,-33.69 71.21,-31.15 80.41,-28.75\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"81.01,-32.21 89.8,-26.29 79.24,-25.44 81.01,-32.21\"/>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M142.98,-25.97C151.51,-28.2 161.21,-30.74 170.41,-33.14\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"169.24,-36.45 179.8,-35.6 171.01,-29.68 169.24,-36.45\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5b0ecd1df0>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[(\"U\", \"Y\"), (\"U\", \"X\"), (\"X\", \"Y\")],\n",
    "    node_props={\n",
    "        \"U\": {\"color\": \"red\"},\n",
    "    },\n",
    "    edge_props={(\"U\", \"X\"): {\"color\": \"red\"}, (\"U\", \"Y\"): {\"color\": \"red\"}},\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4d5ee8fc",
   "metadata": {},
   "source": [
    "In the Fork example, we've shown that stratifying by the confound, we \"close\" the fork by conditioning on U, thus removing any of the causal effect of U on X, thus allowing us to isolate the treatment's effect on Y.\n",
    "\n",
    "This procedure is part of what is known as Do-calculus. The operator `do(X)` tends to mean intervening on X (i.e. setting it to a specific value that is independent of the confound)\n",
    "\n",
    "$$\n",
    "p(Y | do(X)) = \\sum_U p(Y | X, U)p(U) = \\mathbb{E}_U[p(Y | X, U)]\n",
    "$$\n",
    "\n",
    "i.e. **the procedure that gives us the intervention on X is equivalent of the distribution of Y, stratified by the treatment X and the confound X, averaged over the distribution of the confound.**\n",
    "\n",
    "> Note that when we use linear regression estimator for each X, we are **implicity marginalizing and averaging** over out treatment and confound (e.g. in the model form $Y \\sim \\mathcal{N}(\\alpha + \\beta_X X + \\beta_Z Z, \\sigma^2)$\n",
    "\n",
    "- it is generally **not the estimated coefficient** in the model that relate X to Y\n",
    "- it is the distribution of Y when we change X, **averaged over the distribution** defined by the control/confound variables (i.e. U)\n",
    "\n",
    "## Do-calculus\n",
    "- Applied to DAGs, provides a set of rules for identifying $p(Y | do(X))$\n",
    "- **Informs what is possible** before picking functions or distributions\n",
    "- **Justifies graphical analysis**\n",
    "- If do-calculus claims that inference is possible no further special assumptions are required for inference\n",
    "  - often additional assumptions can make the inference even stronger\n",
    "\n",
    "## Backdoor Criterion\n",
    "**Shortcut for applying Do-calculus graphically with your eyeballs.** General rule for finding the minimal sufficient adjustment set of variables to condition on.\n",
    "\n",
    "1. Identify **all** paths connecting treatment X to to outcome Y, including those entering and exiting X (association can be directe/undirected, causation is directed)\n",
    "2. Any of those paths **entering X** are backdoor (non-causal) paths\n",
    "3. Find the **adjustment set** of variables that, once conditioned on, \"closes/blocks\" all the backdoor paths identified in\n",
    "\n",
    "#### Backdoor Criterion Example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "f469310b",
   "metadata": {
    "jupyter": {
     "source_hidden": true
    }
   },
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"332pt\" height=\"71pt\"\n",
       " viewBox=\"0.00 0.00 332.00 70.89\" 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 66.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-66.89 328,-66.89 328,4 -4,4\"/>\n",
       "<!-- U -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>U</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"27\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">U</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"117\" cy=\"-43.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"117\" y=\"-39.9\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- U&#45;&gt;Z -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>U&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M52.51,-26.13C61.12,-28.48 70.97,-31.16 80.3,-33.71\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"79.28,-37.06 89.85,-36.31 81.13,-30.31 79.28,-37.06\"/>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"297\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"297\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- U&#45;&gt;Y -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>U&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M54.23,-17.38C65.26,-16.61 78.25,-15.82 90,-15.45 153.97,-13.38 170.03,-13.38 234,-15.45 241.9,-15.7 250.35,-16.14 258.39,-16.63\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"258.05,-20.11 268.26,-17.28 258.51,-13.13 258.05,-20.11\"/>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"207\" cy=\"-43.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"207\" y=\"-39.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>Z&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M144.4,-43.45C151.89,-43.45 160.18,-43.45 168.2,-43.45\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"168.1,-46.95 178.1,-43.45 168.1,-39.95 168.1,-46.95\"/>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M232.51,-36.76C241.12,-34.41 250.97,-31.73 260.3,-29.18\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"261.13,-32.58 269.85,-26.58 259.28,-25.83 261.13,-32.58\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5b0e1a77d0>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    [(\"U\", \"Z\"), (\"Z\", \"X\"), (\"U\", \"Y\"), (\"X\", \"Y\")],\n",
    "    node_props={\"Z\": {\"color\": \"red\"}, \"U\": {\"style\": \"dashed\"}},\n",
    "    edge_props={\n",
    "        (\"U\", \"Z\"): {\"color\": \"red\"},\n",
    "        (\"Z\", \"X\"): {\"color\": \"red\"},\n",
    "        (\"U\", \"Y\"): {\"color\": \"red\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b563ea62",
   "metadata": {},
   "source": [
    "Backdoor path highlighted in red.\n",
    "\n",
    "- If we could measure $U$ we could just stratify by $U$; however, it is unobserved.\n",
    "- However, we can block the backdoor path by conditioning on $Z$, despite not being able to measure $U$.\n",
    "- This works because $Z$ \"knows\" everything we need to know about association between $X$, $Y$ that is due to the unmeasured confound $U$.\n",
    "\n",
    "#### Resulting graph after stratifying by $Z$\n",
    "- The $U \\rightarrow Z \\rightarrow X$ Pipe has now been broken, disassociateing $X$ from the confound $U$\n",
    "- Note: this doesn't remove the confound's effect on $Y$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "bc2024ac",
   "metadata": {
    "jupyter": {
     "source_hidden": true
    }
   },
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"332pt\" height=\"71pt\"\n",
       " viewBox=\"0.00 0.00 332.00 70.89\" 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 66.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-66.89 328,-66.89 328,4 -4,4\"/>\n",
       "<!-- U -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>U</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"27\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">U</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"117\" cy=\"-43.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"117\" y=\"-39.9\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- U&#45;&gt;Z -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>U&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M52.51,-26.13C61.12,-28.48 70.97,-31.16 80.3,-33.71\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"79.28,-37.06 89.85,-36.31 81.13,-30.31 79.28,-37.06\"/>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"297\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"297\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- U&#45;&gt;Y -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>U&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M54.23,-17.38C65.26,-16.61 78.25,-15.82 90,-15.45 153.97,-13.38 170.03,-13.38 234,-15.45 241.9,-15.7 250.35,-16.14 258.39,-16.63\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"258.05,-20.11 268.26,-17.28 258.51,-13.13 258.05,-20.11\"/>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"207\" cy=\"-43.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"207\" y=\"-39.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>Z&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"none\" d=\"M144.4,-43.45C151.89,-43.45 160.18,-43.45 168.2,-43.45\"/>\n",
       "<polygon fill=\"none\" stroke=\"none\" points=\"168.1,-46.95 178.1,-43.45 168.1,-39.95 168.1,-46.95\"/>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M232.51,-36.76C241.12,-34.41 250.97,-31.73 260.3,-29.18\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"261.13,-32.58 269.85,-26.58 259.28,-25.83 261.13,-32.58\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bd0146810>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    [(\"U\", \"Z\"), (\"Z\", \"X\"), (\"U\", \"Y\"), (\"X\", \"Y\")],\n",
    "    node_props={\"Z\": {\"color\": \"red\"}, \"U\": {\"style\": \"dashed\"}},\n",
    "    edge_props={\n",
    "        (\"U\", \"Z\"): {\"color\": \"red\"},\n",
    "        (\"Z\", \"X\"): {\"color\": \"none\"},\n",
    "        (\"U\", \"Y\"): {\"color\": \"red\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bae86319",
   "metadata": {},
   "source": [
    "### Validate through simulation\n",
    "\n",
    "Here we simulate a situation where Y is caused by X and and an unmeasured confound U that also effects Z and X. (We could also prove mathematically, but simulation is quite confincing as well--for me anyways)\n",
    "\n",
    "$$\n",
    "\\begin{align*}\n",
    "U &\\sim \\text{Bernoulli}(0.5) \\\\\n",
    "Z &\\sim \\text{Normal}(\\beta_{UZ}U, 1) \\\\\n",
    "X &\\sim \\text{Normal}(\\beta_{ZX}Z, 1) \\\\\n",
    "Y &\\sim \\text{Normal}(\\alpha + \\beta_{XY}X + \\beta_{UY}U, 1) \\\\\n",
    "\\end{align*}\n",
    "$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "d17b2ecf",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(123)\n",
    "n_samples = 200\n",
    "\n",
    "alpha = 0\n",
    "beta_XY = 0\n",
    "beta_UY = -1\n",
    "beta_UZ = -1\n",
    "beta_ZX = 1\n",
    "\n",
    "U = stats.bernoulli.rvs(p=0.5, size=n_samples)\n",
    "Z = stats.norm.rvs(loc=beta_UZ * U)\n",
    "X = stats.norm.rvs(loc=beta_ZX * Z)\n",
    "Y = stats.norm.rvs(loc=alpha + beta_XY * X + beta_UY * U)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5e9652b3",
   "metadata": {},
   "source": [
    "### Unstratified (confounded) Model\n",
    "\n",
    "\n",
    "$$\n",
    "\\begin{align*}\n",
    "Y &\\sim \\text{Normal}(\\mu_Y, \\sigma_Y) \\\\\n",
    "\\mu_Y &= \\alpha + \\beta_{XY}X \\\\\n",
    "\\alpha &\\sim \\text{Normal}(0, 1) \\\\\n",
    "\\beta_{XY} &\\sim \\text{Normal}(0, 1) \\\\\n",
    "\\sigma_Y &\\sim \\text{Exponential}(1)\n",
    "\\end{align*}\n",
    "$$"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6b78675d-a08e-4a0e-a6a5-fdd48c8212e4",
   "metadata": {},
   "source": [
    "#### Fit the unstratified model, ignoring Z (and U)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "bb2785d8",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Initializing NUTS using jitter+adapt_diag...\n",
      "Multiprocess sampling (4 chains in 4 jobs)\n",
      "NUTS: [alpha, beta_XY, sigma]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d34db1f335394c6fbf226df184c1d39d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Output()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"></pre>\n"
      ],
      "text/plain": []
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Sampling 4 chains for 1_000 tune and 1_000 draw iterations (4_000 + 4_000 draws total) took 1 seconds.\n"
     ]
    }
   ],
   "source": [
    "with pm.Model() as unstratified_model:\n",
    "    # Priors\n",
    "    alpha_ = pm.Normal(\"alpha\", 0, 1)\n",
    "    beta_XY_ = pm.Normal(\"beta_XY\", 0, 1)\n",
    "    sigma_ = pm.Exponential(\"sigma\", 1)\n",
    "\n",
    "    # Likelihood\n",
    "    mu_ = alpha_ + beta_XY_ * X\n",
    "    Y_ = pm.Normal(\"Y\", mu=mu_, sigma=sigma_, observed=Y)\n",
    "    unstratified_model_inference = pm.sample()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "b4cfb112",
   "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>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>hdi_3%</th>\n",
       "      <th>hdi_97%</th>\n",
       "      <th>mcse_mean</th>\n",
       "      <th>mcse_sd</th>\n",
       "      <th>ess_bulk</th>\n",
       "      <th>ess_tail</th>\n",
       "      <th>r_hat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>alpha</th>\n",
       "      <td>-0.381</td>\n",
       "      <td>0.084</td>\n",
       "      <td>-0.535</td>\n",
       "      <td>-0.219</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3503.0</td>\n",
       "      <td>3034.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>beta_XY</th>\n",
       "      <td>0.148</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.244</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3763.0</td>\n",
       "      <td>3040.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sigma</th>\n",
       "      <td>1.114</td>\n",
       "      <td>0.058</td>\n",
       "      <td>1.015</td>\n",
       "      <td>1.227</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.001</td>\n",
       "      <td>6091.0</td>\n",
       "      <td>3259.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          mean     sd  hdi_3%  hdi_97%  mcse_mean  mcse_sd  ess_bulk  \\\n",
       "alpha   -0.381  0.084  -0.535   -0.219      0.001    0.001    3503.0   \n",
       "beta_XY  0.148  0.055   0.043    0.244      0.001    0.001    3763.0   \n",
       "sigma    1.114  0.058   1.015    1.227      0.001    0.001    6091.0   \n",
       "\n",
       "         ess_tail  r_hat  \n",
       "alpha      3034.0    1.0  \n",
       "beta_XY    3040.0    1.0  \n",
       "sigma      3259.0    1.0  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "az.summary(unstratified_model_inference)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3f742bec",
   "metadata": {},
   "source": [
    "### Stratifying by Z (unconfounded)\n",
    "\n",
    "\n",
    "$$\n",
    "\\begin{align*}\n",
    "Y &\\sim \\text{Normal}(\\mu_Y, \\sigma_Y) \\\\\n",
    "\\mu_Y = &\\alpha + \\beta_{XY}X + \\beta_{Z}Z \\\\\n",
    "\\alpha &\\sim \\text{Normal}(0, 1) \\\\\n",
    "\\beta_{*} &\\sim \\text{Normal}(0, 1) \\\\\n",
    "\\sigma_Y &\\sim \\text{Exponential}(1)\n",
    "\\end{align*}\n",
    "$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "74eabb3d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Initializing NUTS using jitter+adapt_diag...\n",
      "Multiprocess sampling (4 chains in 4 jobs)\n",
      "NUTS: [alpha, beta_XY, beta_Z, sigma]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a521d365c923404893fed0f1c6644011",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Output()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"></pre>\n"
      ],
      "text/plain": []
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Sampling 4 chains for 1_000 tune and 1_000 draw iterations (4_000 + 4_000 draws total) took 1 seconds.\n"
     ]
    }
   ],
   "source": [
    "# Fit the stratified Model\n",
    "with pm.Model() as stratified_model:\n",
    "    # Priors\n",
    "    alpha_ = pm.Normal(\"alpha\", 0, 1)\n",
    "    beta_XY_ = pm.Normal(\"beta_XY\", 0, 1)\n",
    "    beta_Z_ = pm.Normal(\"beta_Z\", 0, 1)\n",
    "    sigma_ = pm.Exponential(\"sigma\", 1)\n",
    "\n",
    "    # Likelihood (includes Z)\n",
    "    mu_ = alpha_ + beta_XY_ * X + beta_Z_ * Z\n",
    "    Y_ = pm.Normal(\"Y\", mu=mu_, sigma=sigma_, observed=Y)\n",
    "    stratified_model_inference = pm.sample()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "5f3f2726",
   "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>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>hdi_3%</th>\n",
       "      <th>hdi_97%</th>\n",
       "      <th>mcse_mean</th>\n",
       "      <th>mcse_sd</th>\n",
       "      <th>ess_bulk</th>\n",
       "      <th>ess_tail</th>\n",
       "      <th>r_hat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>alpha</th>\n",
       "      <td>-0.286</td>\n",
       "      <td>0.091</td>\n",
       "      <td>-0.450</td>\n",
       "      <td>-0.109</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3570.0</td>\n",
       "      <td>2743.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>beta_XY</th>\n",
       "      <td>-0.026</td>\n",
       "      <td>0.078</td>\n",
       "      <td>-0.181</td>\n",
       "      <td>0.114</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.001</td>\n",
       "      <td>2989.0</td>\n",
       "      <td>2631.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>beta_Z</th>\n",
       "      <td>0.318</td>\n",
       "      <td>0.106</td>\n",
       "      <td>0.118</td>\n",
       "      <td>0.515</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.001</td>\n",
       "      <td>2650.0</td>\n",
       "      <td>2469.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sigma</th>\n",
       "      <td>1.091</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.983</td>\n",
       "      <td>1.190</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3813.0</td>\n",
       "      <td>2660.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          mean     sd  hdi_3%  hdi_97%  mcse_mean  mcse_sd  ess_bulk  \\\n",
       "alpha   -0.286  0.091  -0.450   -0.109      0.002    0.001    3570.0   \n",
       "beta_XY -0.026  0.078  -0.181    0.114      0.001    0.001    2989.0   \n",
       "beta_Z   0.318  0.106   0.118    0.515      0.002    0.001    2650.0   \n",
       "sigma    1.091  0.055   0.983    1.190      0.001    0.001    3813.0   \n",
       "\n",
       "         ess_tail  r_hat  \n",
       "alpha      2743.0    1.0  \n",
       "beta_XY    2631.0    1.0  \n",
       "beta_Z     2469.0    1.0  \n",
       "sigma      2660.0    1.0  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "az.summary(stratified_model_inference)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5155001e",
   "metadata": {},
   "source": [
    "**NOTE**: the model coefficient `beta_Z` means nothing in in terms of causal effect of $Z$ on $Y$. In order to determine the causal effect of $Z$ on $Y$ you'd need a different estimator. In general, variables in the adjustment set are not interpretable. This is related to the \"Table 2 Fallacy\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "918a23d4-e27a-4411-92dd-3274bca22efd",
   "metadata": {},
   "source": [
    "#### Compare stratified and unstratified models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "a5fa519a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "_, ax = plt.subplots(figsize=(8, 3))\n",
    "az.plot_dist(unstratified_model_inference.posterior.beta_XY, color=\"k\", label=\"Y|X\", ax=ax)\n",
    "az.plot_dist(stratified_model_inference.posterior.beta_XY, color=\"C0\", label=\"Y|X,Z\", ax=ax)\n",
    "plt.axvline(beta_XY, color=\"C0\", linestyle=\"--\", alpha=0.5, label=f\"Actual={beta_XY}\")\n",
    "plt.xlim([-0.5, 0.5])\n",
    "plt.xlabel(\"posterior beta_XY\")\n",
    "plt.ylabel(\"density\")\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "036b64b3-5fe8-4d39-95b5-3beb8e92fec9",
   "metadata": {},
   "source": [
    "## More Complicated Example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "2fc8aca5-c156-414e-ac58-8b7ca9b4c066",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"332pt\" height=\"139pt\"\n",
       " viewBox=\"0.00 0.00 332.00 139.45\" 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(4 135.45)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-135.45 328,-135.45 328,4 -4,4\"/>\n",
       "<!-- A -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>A</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-112.01\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-108.46\" font-family=\"Times,serif\" font-size=\"14.00\">A</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"117\" cy=\"-89.01\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"117\" y=\"-85.46\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;Z -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>A&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M52.98,-105.48C61.51,-103.25 71.21,-100.71 80.41,-98.31\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"81.01,-101.77 89.8,-95.86 79.24,-95 81.01,-101.77\"/>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"207\" cy=\"-112.01\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"207\" y=\"-108.46\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>A&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M54.22,-114.59C65.25,-115.56 78.24,-116.54 90,-117.01 113.98,-117.97 120.02,-117.97 144,-117.01 151.9,-116.69 160.36,-116.15 168.4,-115.53\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"168.59,-119.02 178.27,-114.72 168.02,-112.05 168.59,-119.02\"/>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>Z&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M142.98,-95.54C151.51,-97.77 161.21,-100.3 170.41,-102.71\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"169.24,-106.02 179.8,-105.16 171.01,-99.24 169.24,-106.02\"/>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"297\" cy=\"-46.01\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"297\" y=\"-42.46\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;Y -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>Z&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M143.13,-82.94C173.6,-75.58 225.21,-63.11 259.93,-54.72\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"260.63,-58.15 269.53,-52.4 258.99,-51.35 260.63,-58.15\"/>\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=\"M226.23,-98.33C238.42,-89.19 254.69,-76.99 268.44,-66.68\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"270.38,-69.6 276.28,-60.8 266.18,-64 270.38,-69.6\"/>\n",
       "</g>\n",
       "<!-- B -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>B</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-32.01\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-28.46\" font-family=\"Times,serif\" font-size=\"14.00\">B</text>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Z -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>B&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M47.53,-44.66C59.07,-52.13 73.9,-61.74 86.81,-70.1\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"84.65,-72.88 94.95,-75.37 88.46,-67 84.65,-72.88\"/>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Y -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>B&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M48.75,-19.96C60.45,-13.87 75.57,-7.12 90,-4.01 151.28,9.19 223.06,-14.23 263.67,-31.07\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"262.05,-34.19 272.62,-34.91 264.81,-27.75 262.05,-34.19\"/>\n",
       "</g>\n",
       "<!-- C -->\n",
       "<g id=\"node6\" class=\"node\">\n",
       "<title>C</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"117\" cy=\"-32.01\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"117\" y=\"-28.46\" font-family=\"Times,serif\" font-size=\"14.00\">C</text>\n",
       "</g>\n",
       "<!-- C&#45;&gt;X -->\n",
       "<g id=\"edge8\" class=\"edge\">\n",
       "<title>C&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M134.55,-47.06C147.7,-59.01 166.3,-75.92 181.23,-89.49\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"178.65,-91.87 188.4,-96.01 183.36,-86.69 178.65,-91.87\"/>\n",
       "</g>\n",
       "<!-- C&#45;&gt;Y -->\n",
       "<g id=\"edge9\" class=\"edge\">\n",
       "<title>C&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M144.26,-34.07C174.46,-36.45 224.24,-40.36 258.5,-43.06\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"258.17,-46.54 268.41,-43.84 258.71,-39.57 258.17,-46.54\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc07e87a0>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"A\", \"Z\"),\n",
    "        (\"A\", \"X\"),\n",
    "        (\"Z\", \"X\"),\n",
    "        (\"B\", \"Z\"),\n",
    "        (\"B\", \"Y\"),\n",
    "        (\"Z\", \"Y\"),\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"C\", \"X\"),\n",
    "        (\"C\", \"Y\"),\n",
    "    ],\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "30a9a524-a318-412c-bfd6-2247103d6da6",
   "metadata": {},
   "source": [
    "### All Paths Connecting X to Y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "7b38d191-7a84-4629-9c89-158ede09ad3f",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"458pt\" height=\"158pt\"\n",
       " viewBox=\"0.00 0.00 457.75 158.12\" 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(4 154.12)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-154.12 453.75,-154.12 453.75,4 -4,4\"/>\n",
       "<!-- A -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>A</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-129.68\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-126.13\" font-family=\"Times,serif\" font-size=\"14.00\">A</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"118\" cy=\"-95.68\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-92.13\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;Z -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>A&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.38,-120.74C61.02,-117.06 72.37,-112.73 82.9,-108.7\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"83.96,-112.05 92.05,-105.21 81.46,-105.51 83.96,-112.05\"/>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"209\" cy=\"-130.68\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"209\" y=\"-127.13\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>A&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M54.17,-129.82C84.75,-129.99 135.57,-130.27 170.37,-130.47\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"169.97,-133.96 179.99,-130.52 170.01,-126.96 169.97,-133.96\"/>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>Z&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M142.38,-104.87C152.09,-108.69 163.55,-113.19 174.15,-117.36\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"172.78,-120.58 183.36,-120.99 175.34,-114.07 172.78,-120.58\"/>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"422.75\" cy=\"-58.68\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"422.75\" y=\"-55.13\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;Y -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>Z&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M145.12,-93.02C191.86,-88.18 292.85,-77.35 377.75,-65.68 380.08,-65.35 382.49,-65.01 384.91,-64.65\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"385.33,-68.13 394.69,-63.16 384.27,-61.21 385.33,-68.13\"/>\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=\"blue\" d=\"M236.36,-129.74C270.6,-127.6 331.47,-120.59 377.75,-97.68 385.42,-93.88 392.79,-88.38 399.23,-82.71\"/>\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"401.42,-85.46 406.28,-76.04 396.61,-80.37 401.42,-85.46\"/>\n",
       "<text text-anchor=\"middle\" x=\"315.88\" y=\"-133.88\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"blue\">Direct Causal Path</text>\n",
       "</g>\n",
       "<!-- B -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>B</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-49.68\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-46.13\" font-family=\"Times,serif\" font-size=\"14.00\">B</text>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Z -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>B&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M49.54,-60.81C60.29,-66.36 73.49,-73.19 85.36,-79.32\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"83.51,-82.31 94,-83.79 86.72,-76.09 83.51,-82.31\"/>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Y -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>B&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M45.46,-35.16C57.44,-26.17 74.17,-15.46 91,-10.68 197.13,19.52 328.09,-21.63 388.43,-44.79\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"386.86,-47.94 397.45,-48.34 389.42,-41.42 386.86,-47.94\"/>\n",
       "</g>\n",
       "<!-- C -->\n",
       "<g id=\"node6\" class=\"node\">\n",
       "<title>C</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"118\" cy=\"-38.68\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-35.13\" font-family=\"Times,serif\" font-size=\"14.00\">C</text>\n",
       "</g>\n",
       "<!-- C&#45;&gt;X -->\n",
       "<g id=\"edge8\" class=\"edge\">\n",
       "<title>C&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M134.1,-54.28C148.03,-68.69 168.94,-90.29 184.96,-106.86\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"182.31,-109.15 191.78,-113.9 187.34,-104.28 182.31,-109.15\"/>\n",
       "</g>\n",
       "<!-- C&#45;&gt;Y -->\n",
       "<g id=\"edge9\" class=\"edge\">\n",
       "<title>C&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M145.34,-40.42C199.26,-43.98 322.38,-52.11 384.48,-56.21\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"383.97,-59.69 394.18,-56.85 384.43,-52.7 383.97,-59.69\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc09c3140>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"A\", \"Z\"),\n",
    "        (\"A\", \"X\"),\n",
    "        (\"Z\", \"X\"),\n",
    "        (\"B\", \"Z\"),\n",
    "        (\"B\", \"Y\"),\n",
    "        (\"Z\", \"Y\"),\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"C\", \"X\"),\n",
    "        (\"C\", \"Y\"),\n",
    "    ],\n",
    "    edge_props={\n",
    "        (\"X\", \"Y\"): {\"color\": \"blue\", \"label\": \"Direct Causal Path\", \"fontcolor\": \"blue\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "873fe316-98ea-41bf-b97b-44759afe8ea3",
   "metadata": {},
   "source": [
    "#### $X \\rightarrow Y$\n",
    "- Direct, causal path, leave open"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "7d6f562e-f610-40e6-99cf-b812c2ce86c2",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"360pt\" height=\"166pt\"\n",
       " viewBox=\"0.00 0.00 360.25 165.78\" 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(4 161.78)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-161.78 356.25,-161.78 356.25,4 -4,4\"/>\n",
       "<!-- A -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>A</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-127.34\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-123.79\" font-family=\"Times,serif\" font-size=\"14.00\">A</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"118\" cy=\"-93.34\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-89.79\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;Z -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>A&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.38,-118.4C61.02,-114.72 72.37,-110.39 82.9,-106.37\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"83.96,-109.71 92.05,-102.87 81.46,-103.17 83.96,-109.71\"/>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"234.25\" cy=\"-138.34\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"234.25\" y=\"-134.79\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>A&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M54.41,-128.75C90.28,-130.67 154.61,-134.12 195.55,-136.32\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"195.25,-139.81 205.43,-136.85 195.63,-132.82 195.25,-139.81\"/>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>Z&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M142.12,-102.44C158.61,-108.94 181.05,-117.78 199.53,-125.06\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"198.17,-128.28 208.76,-128.69 200.73,-121.77 198.17,-128.28\"/>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"325.25\" cy=\"-57.34\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"325.25\" y=\"-53.79\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;Y -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>Z&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M144.78,-89.7C176.69,-85.07 232.65,-76.57 280.25,-67.34 282.85,-66.83 285.52,-66.29 288.21,-65.72\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"288.68,-69.2 297.7,-63.65 287.18,-62.36 288.68,-69.2\"/>\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=\"M254.72,-125.32C262.89,-119.54 272.31,-112.45 280.25,-105.34 288.2,-98.21 296.27,-89.81 303.3,-82.04\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"305.72,-84.59 309.72,-74.78 300.47,-79.95 305.72,-84.59\"/>\n",
       "</g>\n",
       "<!-- B -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>B</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-44.34\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-40.79\" font-family=\"Times,serif\" font-size=\"14.00\">B</text>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Z -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>B&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M49.09,-55.95C60.1,-62.01 73.77,-69.53 85.96,-76.24\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"83.98,-79.15 94.43,-80.91 87.35,-73.02 83.98,-79.15\"/>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Y -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>B&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M45.78,-29.99C57.78,-21.27 74.39,-10.96 91,-6.34 162.39,13.53 247.81,-19.22 292.86,-40.79\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"291.27,-43.91 301.79,-45.19 294.37,-37.63 291.27,-43.91\"/>\n",
       "</g>\n",
       "<!-- C -->\n",
       "<g id=\"node6\" class=\"node\">\n",
       "<title>C</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"118\" cy=\"-34.34\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-30.79\" font-family=\"Times,serif\" font-size=\"14.00\">C</text>\n",
       "</g>\n",
       "<!-- C&#45;&gt;X -->\n",
       "<g id=\"edge8\" class=\"edge\">\n",
       "<title>C&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M138.32,-47.43C152.92,-57.71 173.13,-72.75 189.25,-87.84 197.48,-95.54 205.78,-104.65 212.93,-113\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"210.1,-115.07 219.2,-120.49 215.47,-110.58 210.1,-115.07\"/>\n",
       "<text text-anchor=\"middle\" x=\"176.12\" y=\"-93.54\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"red\">XCY</text>\n",
       "</g>\n",
       "<!-- C&#45;&gt;Y -->\n",
       "<g id=\"edge9\" class=\"edge\">\n",
       "<title>C&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M145,-37.25C180.83,-41.27 245.63,-48.53 286.72,-53.13\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"286.31,-56.61 296.64,-54.24 287.09,-49.65 286.31,-56.61\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0971850>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"A\", \"Z\"),\n",
    "        (\"A\", \"X\"),\n",
    "        (\"Z\", \"X\"),\n",
    "        (\"B\", \"Z\"),\n",
    "        (\"B\", \"Y\"),\n",
    "        (\"Z\", \"Y\"),\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"C\", \"X\"),\n",
    "        (\"C\", \"Y\"),\n",
    "    ],\n",
    "    edge_props={\n",
    "        (\"C\", \"X\"): {\"color\": \"red\", \"label\": \"XCY\", \"fontcolor\": \"red\"},\n",
    "        (\"C\", \"Y\"): {\"color\": \"red\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9d9e70ce-0043-42e0-8337-ffe28f9497f0",
   "metadata": {},
   "source": [
    "#### $X \\leftarrow C \\rightarrow Y$\n",
    "- Backdoor non-causal path\n",
    "- Block by **stratifying by $C$**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "410970d8-34e7-4852-b2df-c58d062da5d8",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"360pt\" height=\"155pt\"\n",
       " viewBox=\"0.00 0.00 360.25 155.20\" 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(4 151.2)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-151.2 356.25,-151.2 356.25,4 -4,4\"/>\n",
       "<!-- A -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>A</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-127.76\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-124.21\" font-family=\"Times,serif\" font-size=\"14.00\">A</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"118\" cy=\"-93.76\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-90.21\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;Z -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>A&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.38,-118.82C61.02,-115.14 72.37,-110.81 82.9,-106.79\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"83.96,-110.13 92.05,-103.29 81.46,-103.59 83.96,-110.13\"/>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"234.25\" cy=\"-120.76\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"234.25\" y=\"-117.21\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>A&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M54.3,-129.03C86.27,-130.28 141.81,-131.51 189.25,-127.76 191.6,-127.57 194.01,-127.33 196.44,-127.05\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"196.8,-130.53 206.24,-125.72 195.86,-123.6 196.8,-130.53\"/>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>Z&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M144.46,-98.55C157.85,-101.18 174.49,-104.64 189.25,-108.26 192.19,-108.98 195.22,-109.76 198.26,-110.57\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"197.1,-113.88 207.67,-113.16 198.96,-107.13 197.1,-113.88\"/>\n",
       "<text text-anchor=\"middle\" x=\"176.12\" y=\"-113.96\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"red\">XZY</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"325.25\" cy=\"-48.76\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"325.25\" y=\"-45.21\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;Y -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>Z&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M144.48,-88.69C176.32,-82.29 232.44,-70.77 280.25,-59.76 283,-59.12 285.85,-58.45 288.71,-57.76\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"289.27,-61.23 298.15,-55.45 287.61,-54.43 289.27,-61.23\"/>\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=\"M254.37,-107.33C262.48,-101.48 271.96,-94.45 280.25,-87.76 286.55,-82.67 293.16,-77 299.3,-71.59\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"301.6,-74.23 306.72,-64.95 296.93,-69.01 301.6,-74.23\"/>\n",
       "</g>\n",
       "<!-- B -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>B</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-44.76\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-41.21\" font-family=\"Times,serif\" font-size=\"14.00\">B</text>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Z -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>B&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M49.09,-56.37C60.1,-62.43 73.77,-69.95 85.96,-76.66\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"83.98,-79.57 94.43,-81.33 87.35,-73.44 83.98,-79.57\"/>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Y -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>B&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M45.78,-30.41C57.78,-21.69 74.39,-11.38 91,-6.76 161.31,12.81 246.17,-14.9 291.66,-33.77\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"290.11,-36.92 300.68,-37.63 292.86,-30.48 290.11,-36.92\"/>\n",
       "</g>\n",
       "<!-- C -->\n",
       "<g id=\"node6\" class=\"node\">\n",
       "<title>C</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"118\" cy=\"-34.76\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-31.21\" font-family=\"Times,serif\" font-size=\"14.00\">C</text>\n",
       "</g>\n",
       "<!-- C&#45;&gt;X -->\n",
       "<g id=\"edge8\" class=\"edge\">\n",
       "<title>C&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M137.36,-48.57C155.84,-62.49 184.59,-84.13 205.79,-100.09\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"203.64,-102.85 213.73,-106.06 207.85,-97.25 203.64,-102.85\"/>\n",
       "</g>\n",
       "<!-- C&#45;&gt;Y -->\n",
       "<g id=\"edge9\" class=\"edge\">\n",
       "<title>C&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M145,-36.53C180.83,-38.97 245.63,-43.39 286.72,-46.2\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"286.42,-49.68 296.63,-46.87 286.89,-42.7 286.42,-49.68\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc09c1d90>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"A\", \"Z\"),\n",
    "        (\"A\", \"X\"),\n",
    "        (\"Z\", \"X\"),\n",
    "        (\"B\", \"Z\"),\n",
    "        (\"B\", \"Y\"),\n",
    "        (\"Z\", \"Y\"),\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"C\", \"X\"),\n",
    "        (\"C\", \"Y\"),\n",
    "    ],\n",
    "    edge_props={\n",
    "        (\"Z\", \"X\"): {\"color\": \"red\", \"label\": \"XZY\", \"fontcolor\": \"red\"},\n",
    "        (\"Z\", \"Y\"): {\"color\": \"red\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e0a7b311-472c-49b2-a282-55a42d58cb72",
   "metadata": {},
   "source": [
    "#### $X \\leftarrow Z \\rightarrow Y$\n",
    "- Backdoor non-causal path\n",
    "- Block by **stratifying by $Z$**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "8365fbb9-8408-4a4a-9402-3ccc0c7a149f",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"335pt\" height=\"160pt\"\n",
       " viewBox=\"0.00 0.00 335.00 160.23\" 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(4 156.23)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-156.23 331,-156.23 331,4 -4,4\"/>\n",
       "<!-- A -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>A</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-132.79\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-129.24\" font-family=\"Times,serif\" font-size=\"14.00\">A</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"118\" cy=\"-91.79\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-88.24\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;Z -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>A&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M50.46,-122.44C60.68,-117.73 72.98,-112.07 84.21,-106.89\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"85.56,-110.13 93.17,-102.76 82.63,-103.77 85.56,-110.13\"/>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"209\" cy=\"-126.79\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"209\" y=\"-123.24\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>A&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M54.17,-131.92C84.83,-130.89 135.86,-129.19 170.66,-128.03\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"170.4,-131.54 180.28,-127.71 170.17,-124.55 170.4,-131.54\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-135.99\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"red\">XAZBY</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>Z&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M142.38,-100.98C152.09,-104.8 163.55,-109.31 174.15,-113.48\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"172.78,-116.7 183.36,-117.1 175.34,-110.18 172.78,-116.7\"/>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"300\" cy=\"-54.79\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"300\" y=\"-51.24\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;Y -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>Z&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M144.61,-86.92C171.98,-81.67 216.62,-72.96 255,-64.79 257.59,-64.24 260.25,-63.66 262.94,-63.07\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"263.42,-66.55 272.42,-60.95 261.9,-59.72 263.42,-66.55\"/>\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=\"M229.12,-113.36C237.23,-107.52 246.71,-100.48 255,-93.79 261.3,-88.71 267.91,-83.03 274.05,-77.62\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"276.35,-80.26 281.47,-70.98 271.68,-75.04 276.35,-80.26\"/>\n",
       "</g>\n",
       "<!-- B -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>B</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-52.79\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-49.24\" font-family=\"Times,serif\" font-size=\"14.00\">B</text>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Z -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>B&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M50.46,-62.63C60.6,-67.08 72.8,-72.42 83.96,-77.31\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"82.29,-80.4 92.85,-81.21 85.1,-73.99 82.29,-80.4\"/>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Y -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>B&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M43.48,-37.1C55.33,-26.25 72.72,-12.75 91,-6.79 152.61,13.31 227.11,-16.45 268.14,-37.33\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"266.48,-40.41 276.96,-41.97 269.74,-34.21 266.48,-40.41\"/>\n",
       "</g>\n",
       "<!-- C -->\n",
       "<g id=\"node6\" class=\"node\">\n",
       "<title>C</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"118\" cy=\"-34.79\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-31.24\" font-family=\"Times,serif\" font-size=\"14.00\">C</text>\n",
       "</g>\n",
       "<!-- C&#45;&gt;X -->\n",
       "<g id=\"edge8\" class=\"edge\">\n",
       "<title>C&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M134.1,-50.4C148.03,-64.8 168.94,-86.41 184.96,-102.97\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"182.31,-105.26 191.78,-110.02 187.34,-100.4 182.31,-105.26\"/>\n",
       "</g>\n",
       "<!-- C&#45;&gt;Y -->\n",
       "<g id=\"edge9\" class=\"edge\">\n",
       "<title>C&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M145.17,-37.7C175.83,-41.1 226.86,-46.77 261.66,-50.64\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"260.97,-54.08 271.29,-51.71 261.74,-47.13 260.97,-54.08\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc095ad80>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"A\", \"Z\"),\n",
    "        (\"A\", \"X\"),\n",
    "        (\"Z\", \"X\"),\n",
    "        (\"B\", \"Z\"),\n",
    "        (\"B\", \"Y\"),\n",
    "        (\"Z\", \"Y\"),\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"C\", \"X\"),\n",
    "        (\"C\", \"Y\"),\n",
    "    ],\n",
    "    edge_props={\n",
    "        (\"A\", \"X\"): {\"color\": \"red\", \"label\": \"XAZBY\", \"fontcolor\": \"red\"},\n",
    "        (\"A\", \"Z\"): {\"color\": \"red\"},\n",
    "        (\"B\", \"Z\"): {\"color\": \"red\"},\n",
    "        (\"B\", \"Y\"): {\"color\": \"red\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1abf64e5-f16a-4498-aa82-835e57535bb9",
   "metadata": {},
   "source": [
    "#### $X \\leftarrow A \\rightarrow Z \\leftarrow B \\rightarrow Y$\n",
    "- Backdoor non-causal path\n",
    "- Block by **stratifying by $A$ or $B$**; stratifying by $Z$ opens the path b.c. it's a collider\n",
    "  - we're already stratifying by $Z$ for the $X,Z,Y$ backdoor path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "86e219fb-de71-448c-b647-5b757dfd436c",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"335pt\" height=\"160pt\"\n",
       " viewBox=\"0.00 0.00 335.00 160.23\" 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(4 156.23)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-156.23 331,-156.23 331,4 -4,4\"/>\n",
       "<!-- A -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>A</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-132.79\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-129.24\" font-family=\"Times,serif\" font-size=\"14.00\">A</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"118\" cy=\"-91.79\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-88.24\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;Z -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>A&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M50.46,-122.44C60.68,-117.73 72.98,-112.07 84.21,-106.89\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"85.56,-110.13 93.17,-102.76 82.63,-103.77 85.56,-110.13\"/>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"209\" cy=\"-126.79\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"209\" y=\"-123.24\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>A&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M54.17,-131.92C84.83,-130.89 135.86,-129.19 170.66,-128.03\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"170.4,-131.54 180.28,-127.71 170.17,-124.55 170.4,-131.54\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-135.99\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"red\">XAZY</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>Z&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M142.38,-100.98C152.09,-104.8 163.55,-109.31 174.15,-113.48\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"172.78,-116.7 183.36,-117.1 175.34,-110.18 172.78,-116.7\"/>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"300\" cy=\"-54.79\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"300\" y=\"-51.24\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;Y -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>Z&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M144.61,-86.92C171.98,-81.67 216.62,-72.96 255,-64.79 257.59,-64.24 260.25,-63.66 262.94,-63.07\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"263.42,-66.55 272.42,-60.95 261.9,-59.72 263.42,-66.55\"/>\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=\"M229.12,-113.36C237.23,-107.52 246.71,-100.48 255,-93.79 261.3,-88.71 267.91,-83.03 274.05,-77.62\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"276.35,-80.26 281.47,-70.98 271.68,-75.04 276.35,-80.26\"/>\n",
       "</g>\n",
       "<!-- B -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>B</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-52.79\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-49.24\" font-family=\"Times,serif\" font-size=\"14.00\">B</text>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Z -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>B&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M50.46,-62.63C60.6,-67.08 72.8,-72.42 83.96,-77.31\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"82.29,-80.4 92.85,-81.21 85.1,-73.99 82.29,-80.4\"/>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Y -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>B&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M43.48,-37.1C55.33,-26.25 72.72,-12.75 91,-6.79 152.61,13.31 227.11,-16.45 268.14,-37.33\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"266.48,-40.41 276.96,-41.97 269.74,-34.21 266.48,-40.41\"/>\n",
       "</g>\n",
       "<!-- C -->\n",
       "<g id=\"node6\" class=\"node\">\n",
       "<title>C</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"118\" cy=\"-34.79\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-31.24\" font-family=\"Times,serif\" font-size=\"14.00\">C</text>\n",
       "</g>\n",
       "<!-- C&#45;&gt;X -->\n",
       "<g id=\"edge8\" class=\"edge\">\n",
       "<title>C&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M134.1,-50.4C148.03,-64.8 168.94,-86.41 184.96,-102.97\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"182.31,-105.26 191.78,-110.02 187.34,-100.4 182.31,-105.26\"/>\n",
       "</g>\n",
       "<!-- C&#45;&gt;Y -->\n",
       "<g id=\"edge9\" class=\"edge\">\n",
       "<title>C&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M145.17,-37.7C175.83,-41.1 226.86,-46.77 261.66,-50.64\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"260.97,-54.08 271.29,-51.71 261.74,-47.13 260.97,-54.08\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc07ea7e0>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"A\", \"Z\"),\n",
    "        (\"A\", \"X\"),\n",
    "        (\"Z\", \"X\"),\n",
    "        (\"B\", \"Z\"),\n",
    "        (\"B\", \"Y\"),\n",
    "        (\"Z\", \"Y\"),\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"C\", \"X\"),\n",
    "        (\"C\", \"Y\"),\n",
    "    ],\n",
    "    edge_props={\n",
    "        (\"A\", \"X\"): {\"color\": \"red\", \"label\": \"XAZY\", \"fontcolor\": \"red\"},\n",
    "        (\"A\", \"Z\"): {\"color\": \"red\"},\n",
    "        (\"Z\", \"Y\"): {\"color\": \"red\"},\n",
    "        (\"Z\", \"Y\"): {\"color\": \"red\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "888b885b-c066-4d31-888d-38fb706d418a",
   "metadata": {},
   "source": [
    "#### $X \\leftarrow A \\rightarrow Z \\rightarrow Y$\n",
    "- Backdoor non-causal path\n",
    "- Block by **stratifying by $A$;** stratifying by $Z$ opens the path b.c. it's a collider\n",
    "  - we're already stratifying by $Z$ for the $X \\leftarrow Z \\rightarrow Y$ backdoor path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "69def1ed-9f34-4064-932c-5f5e3a55ee01",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"369pt\" height=\"156pt\"\n",
       " viewBox=\"0.00 0.00 368.50 155.53\" 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(4 151.53)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-151.53 364.5,-151.53 364.5,4 -4,4\"/>\n",
       "<!-- A -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>A</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-128.08\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-124.53\" font-family=\"Times,serif\" font-size=\"14.00\">A</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"118\" cy=\"-94.08\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-90.53\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;Z -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>A&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M51.38,-119.15C61.02,-115.47 72.37,-111.13 82.9,-107.11\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"83.96,-110.45 92.05,-103.61 81.46,-103.91 83.96,-110.45\"/>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"242.5\" cy=\"-121.08\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"242.5\" y=\"-117.53\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>A&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M54.18,-129.36C87.53,-130.69 146.9,-132.08 197.5,-128.08 199.85,-127.9 202.26,-127.65 204.69,-127.37\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"205.05,-130.86 214.49,-126.04 204.11,-123.92 205.05,-130.86\"/>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>Z&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M144.58,-98.18C159.99,-100.8 179.96,-104.48 197.5,-108.58 200.44,-109.27 203.48,-110.03 206.53,-110.82\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"205.37,-114.14 215.94,-113.39 207.21,-107.38 205.37,-114.14\"/>\n",
       "<text text-anchor=\"middle\" x=\"180.25\" y=\"-114.28\" font-family=\"Times,serif\" font-size=\"14.00\" fill=\"red\">XZBY</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"333.5\" cy=\"-49.08\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"333.5\" y=\"-45.53\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;Y -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>Z&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M144.64,-89.33C177.9,-83.07 237.7,-71.52 288.5,-60.08 291.26,-59.46 294.1,-58.8 296.96,-58.12\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"297.53,-61.58 306.41,-55.81 295.87,-54.78 297.53,-61.58\"/>\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=\"M262.62,-107.65C270.73,-101.81 280.21,-94.77 288.5,-88.08 294.8,-83 301.41,-77.32 307.55,-71.91\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"309.85,-74.55 314.97,-65.28 305.18,-69.33 309.85,-74.55\"/>\n",
       "</g>\n",
       "<!-- B -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>B</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-45.08\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-41.53\" font-family=\"Times,serif\" font-size=\"14.00\">B</text>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Z -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>B&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M49.09,-56.69C60.1,-62.75 73.77,-70.28 85.96,-76.99\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"83.98,-79.9 94.43,-81.65 87.35,-73.76 83.98,-79.9\"/>\n",
       "</g>\n",
       "<!-- B&#45;&gt;Y -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>B&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M45.78,-30.74C57.78,-22.02 74.39,-11.7 91,-7.08 164.26,13.3 252.9,-15.23 299.76,-34.32\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"298.23,-37.47 308.8,-38.12 300.94,-31.02 298.23,-37.47\"/>\n",
       "</g>\n",
       "<!-- C -->\n",
       "<g id=\"node6\" class=\"node\">\n",
       "<title>C</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"118\" cy=\"-35.08\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-31.53\" font-family=\"Times,serif\" font-size=\"14.00\">C</text>\n",
       "</g>\n",
       "<!-- C&#45;&gt;X -->\n",
       "<g id=\"edge8\" class=\"edge\">\n",
       "<title>C&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M138.14,-48.52C158.28,-62.66 190.18,-85.05 213.22,-101.23\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"211.06,-103.99 221.26,-106.87 215.08,-98.26 211.06,-103.99\"/>\n",
       "</g>\n",
       "<!-- C&#45;&gt;Y -->\n",
       "<g id=\"edge9\" class=\"edge\">\n",
       "<title>C&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M145.18,-36.8C182.63,-39.25 251.73,-43.78 294.72,-46.6\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"294.39,-50.09 304.6,-47.25 294.85,-43.1 294.39,-50.09\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc09717f0>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"A\", \"Z\"),\n",
    "        (\"A\", \"X\"),\n",
    "        (\"Z\", \"X\"),\n",
    "        (\"B\", \"Z\"),\n",
    "        (\"B\", \"Y\"),\n",
    "        (\"Z\", \"Y\"),\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"C\", \"X\"),\n",
    "        (\"C\", \"Y\"),\n",
    "    ],\n",
    "    edge_props={\n",
    "        (\"Z\", \"X\"): {\"color\": \"red\", \"label\": \"XZBY\", \"fontcolor\": \"red\"},\n",
    "        (\"B\", \"Z\"): {\"color\": \"red\"},\n",
    "        (\"B\", \"Y\"): {\"color\": \"red\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "90879f15-aa17-4b51-8eda-c1071787cbe5",
   "metadata": {},
   "source": [
    "#### $X \\leftarrow Z \\leftarrow B \\rightarrow Y$\n",
    "- Backdoor non-causal path\n",
    "- Block by **stratifying by $B$**; stratifying by $Z$ opens the path b.c. it's a collider\n",
    "  - we're already stratifying by $Z$ for the $X \\leftarrow Z \\rightarrow Y$ backdoor path"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f63c5a3c-1706-47b9-8cb4-75dc378fa308",
   "metadata": {},
   "source": [
    "#### Resulting Minimum Adjustment set: Z, C, (A or B)\n",
    "Chossing B over A turns out to be more statistically efficient, though not causally different than choosing A"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "63944882-aa94-43eb-8329-68bdcf55714e",
   "metadata": {},
   "source": [
    "## Example with unobserved confounds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "b085c2f6-fbd2-4b97-b219-3e90ec7ecc33",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"382pt\" height=\"170pt\"\n",
       " viewBox=\"0.00 0.00 381.84 169.89\" 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(4 165.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-165.89 377.84,-165.89 377.84,4 -4,4\"/>\n",
       "<!-- G -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>G</title>\n",
       "<ellipse fill=\"none\" stroke=\"blue\" cx=\"67\" cy=\"-85.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-81.9\" font-family=\"Times,serif\" font-size=\"14.00\">G</text>\n",
       "</g>\n",
       "<!-- P -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>P</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"226.92\" cy=\"-48.45\" rx=\"56.92\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"226.92\" y=\"-44.9\" font-family=\"Times,serif\" font-size=\"14.00\">collider, P</text>\n",
       "</g>\n",
       "<!-- G&#45;&gt;P -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>G&#45;&gt;P</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M93.02,-79.58C113.1,-74.88 142.09,-68.09 167.87,-62.05\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"168.65,-65.46 177.59,-59.77 167.06,-58.64 168.65,-65.46\"/>\n",
       "</g>\n",
       "<!-- C -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>C</title>\n",
       "<ellipse fill=\"none\" stroke=\"blue\" cx=\"346.84\" cy=\"-48.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"346.84\" y=\"-44.9\" font-family=\"Times,serif\" font-size=\"14.00\">C</text>\n",
       "</g>\n",
       "<!-- G&#45;&gt;C -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>G&#45;&gt;C</title>\n",
       "<path fill=\"none\" stroke=\"blue\" d=\"M94.02,-87.51C135.01,-90.07 216.88,-92.33 283.84,-76.45 294.28,-73.97 305.11,-69.73 314.75,-65.28\"/>\n",
       "<polygon fill=\"blue\" stroke=\"blue\" points=\"316.22,-68.46 323.68,-60.92 313.15,-62.17 316.22,-68.46\"/>\n",
       "</g>\n",
       "<!-- P&#45;&gt;C -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>P&#45;&gt;C</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M284.12,-48.45C292.21,-48.45 300.38,-48.45 308,-48.45\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"307.94,-51.95 317.94,-48.45 307.94,-44.95 307.94,-51.95\"/>\n",
       "</g>\n",
       "<!-- u -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>u</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"67\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">u</text>\n",
       "</g>\n",
       "<!-- u&#45;&gt;P -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>u&#45;&gt;P</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M93.37,-24.1C112.65,-27.65 140,-32.67 164.86,-37.23\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"164.11,-40.65 174.57,-39.02 165.37,-33.77 164.11,-40.65\"/>\n",
       "</g>\n",
       "<!-- u&#45;&gt;C -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>u&#45;&gt;C</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M93.96,-16.12C134.89,-11.64 216.65,-5.81 283.84,-20.45 294.32,-22.73 305.17,-26.91 314.8,-31.37\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"313.21,-34.48 323.73,-35.75 316.3,-28.2 313.21,-34.48\"/>\n",
       "</g>\n",
       "<!-- unobserved -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>unobserved</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"67\" cy=\"-142.45\" rx=\"67\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-138.9\" font-family=\"Times,serif\" font-size=\"14.00\">unobserved</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc095b920>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[(\"G\", \"P\"), (\"P\", \"C\"), (\"G\", \"C\"), (\"u\", \"P\"), (\"u\", \"C\")],\n",
    "    node_props={\n",
    "        \"G\": {\"color\": \"blue\"},\n",
    "        \"C\": {\"color\": \"blue\"},\n",
    "        \"P\": {\"color\": \"red\", \"label\": \"collider, P\"},\n",
    "        \"u\": {\"style\": \"dashed\"},\n",
    "        \"unobserved\": {\"style\": \"dashed\"},\n",
    "    },\n",
    "    edge_props={(\"G\", \"C\"): {\"color\": \"blue\"}},\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b5e37c20-b75d-4144-89d5-cae0b99c2853",
   "metadata": {},
   "source": [
    "- $P$ is mediator between $G$ and $C$\n",
    "- $P$ is also a collider between $G, u, C$\n",
    "- If we want to estimate direct effect of $G \\rightarrow C$, we'll need to stratify by $P$--close the Pipe\n",
    "- However, this will open up the Collider path to the unobserved confound.\n",
    "- **It's not possible to accurately estimate the _Direct Causal Effect_ of G on C**\n",
    "- It **is possible to estimate the _Total Causal Effect_**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4e9ab25e",
   "metadata": {},
   "source": [
    "# Good and bad controls\n",
    "\n",
    "## Common _incorrect_ heuristics for choosing control variables:\n",
    "\n",
    "- YOLO approach -- anything in the spreadsheet\n",
    "- Ignore highly colinear variables\n",
    "  - false, no support for this\n",
    "  - colinearity can arise through many different causal processes that _can be modeled accurately_\n",
    "- It's safe to add pre-treatment variables\n",
    "  - false, pre-treatment, just like post-treatment variables can cause confounds."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f95175c1",
   "metadata": {},
   "source": [
    "## Good & Bad Controls Examples"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3fdbe678-4677-4e72-97a3-2ed2fa5a531c",
   "metadata": {},
   "source": [
    "### Bad control"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "afaf1af1",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"313pt\" height=\"197pt\"\n",
       " viewBox=\"0.00 0.00 313.00 196.67\" 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(4 192.67)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-192.67 309,-192.67 309,4 -4,4\"/>\n",
       "<!-- u -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>u</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"54\" cy=\"-169.23\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"54\" y=\"-165.68\" font-family=\"Times,serif\" font-size=\"14.00\">u</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"99\" cy=\"-94.34\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"99\" y=\"-90.79\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- u&#45;&gt;Z -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>u&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M64.66,-150.95C69.99,-142.33 76.55,-131.7 82.5,-122.06\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"85.34,-124.12 87.62,-113.77 79.39,-120.44 85.34,-124.12\"/>\n",
       "</g>\n",
       "<!-- X -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"27\" cy=\"-94.34\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-90.79\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- u&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>u&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M47.33,-150.21C44.43,-142.38 40.96,-133.02 37.71,-124.26\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"41.07,-123.23 34.31,-115.07 34.5,-125.67 41.07,-123.23\"/>\n",
       "</g>\n",
       "<!-- v -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>v</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"126\" cy=\"-169.23\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"126\" y=\"-165.68\" font-family=\"Times,serif\" font-size=\"14.00\">v</text>\n",
       "</g>\n",
       "<!-- v&#45;&gt;Z -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>v&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M119.33,-150.21C116.43,-142.38 112.96,-133.02 109.71,-124.26\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"113.07,-123.23 106.31,-115.07 106.5,-125.67 113.07,-123.23\"/>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"90\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"90\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- v&#45;&gt;Y -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>v&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M132.26,-150.11C137.88,-130.87 143.98,-99.86 135,-74.89 130.72,-63 122.59,-51.86 114.41,-42.79\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"117.19,-40.63 107.72,-35.86 112.15,-45.49 117.19,-40.63\"/>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M40.67,-77.52C48.93,-67.97 59.62,-55.6 68.94,-44.82\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"71.48,-47.23 75.37,-37.38 66.18,-42.65 71.48,-47.23\"/>\n",
       "</g>\n",
       "<!-- unobserved -->\n",
       "<g id=\"node6\" class=\"node\">\n",
       "<title>unobserved</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"238\" cy=\"-169.23\" rx=\"67\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"238\" y=\"-165.68\" font-family=\"Times,serif\" font-size=\"14.00\">unobserved</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0958e60>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[(\"u\", \"Z\"), (\"v\", \"Z\"), (\"u\", \"X\"), (\"X\", \"Y\"), (\"v\", \"Y\")],\n",
    "    node_props={\n",
    "        \"u\": {\"style\": \"dashed\"},\n",
    "        \"v\": {\"style\": \"dashed\"},\n",
    "        \"X\": {\"color\": \"red\"},\n",
    "        \"Y\": {\"color\": \"red\"},\n",
    "        \"unobserved\": {\"style\": \"dashed\"},\n",
    "    },\n",
    "    edge_props={(\"X\", \"Y\"): {\"color\": \"red\"}},\n",
    "    graph_direction=\"TD\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c9ef216b",
   "metadata": {},
   "source": [
    "$Z$ is a collider for unobserved variables $u$ and $v$, which independently affect $X$ and $Y$\n",
    "#### List the paths\n",
    "- $X \\rightarrow Y$\n",
    "  - causal, leave open\n",
    "- $X \\leftarrow u \\rightarrow Z \\leftarrow v \\rightarrow Y$\n",
    "  - backdoor, closed due to collider\n",
    "  - $Z$ is a **bad control**: stratifying by $Z$ would open the backdoor path\n",
    "  - $Z$ could be a pre-treatment variable -- not always good to stratify by pre-treatment variables; **draw your causal assumptions**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f156a2cf",
   "metadata": {},
   "source": [
    "### Bad mediator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "00a4ed24",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"322pt\" height=\"161pt\"\n",
       " viewBox=\"0.00 0.00 322.00 160.89\" 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(4 156.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-156.89 318,-156.89 318,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"67\" cy=\"-76.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-72.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"197\" cy=\"-66.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"197\" y=\"-62.9\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Z -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M94.21,-74.4C112.76,-72.95 137.91,-70.98 158.57,-69.37\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"158.61,-72.88 168.31,-68.61 158.07,-65.9 158.61,-72.88\"/>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"287\" cy=\"-42.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"287\" y=\"-38.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;Y -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>Z&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M222.51,-59.76C231.11,-57.41 240.97,-54.73 250.3,-52.18\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"251.12,-55.58 259.85,-49.58 249.28,-48.83 251.12,-55.58\"/>\n",
       "</g>\n",
       "<!-- u -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>u</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"67\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">u</text>\n",
       "</g>\n",
       "<!-- u&#45;&gt;Z -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>u&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M91.48,-28.07C111.23,-35.32 139.74,-45.79 162,-53.96\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"160.65,-57.2 171.25,-57.36 163.07,-50.63 160.65,-57.2\"/>\n",
       "</g>\n",
       "<!-- u&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>u&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M93.86,-22.17C132.19,-26.22 204.29,-33.82 248.44,-38.48\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"248.01,-41.96 258.32,-39.53 248.75,-35 248.01,-41.96\"/>\n",
       "</g>\n",
       "<!-- unobserved -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>unobserved</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"67\" cy=\"-133.45\" rx=\"67\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-129.9\" font-family=\"Times,serif\" font-size=\"14.00\">unobserved</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc09730e0>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Z\"),\n",
    "        (\"Z\", \"Y\"),\n",
    "        (\"u\", \"Z\"),\n",
    "        (\"u\", \"Y\"),\n",
    "    ],\n",
    "    node_props={\n",
    "        \"X\": {\"color\": \"red\"},\n",
    "        \"Y\": {\"color\": \"red\"},\n",
    "        \"u\": {\"style\": \"dashed\"},\n",
    "        \"unobserved\": {\"style\": \"dashed\"},\n",
    "    },\n",
    "    edge_props={(\"X\", \"Z\"): {\"color\": \"red\"}, (\"Z\", \"Y\"): {\"color\": \"red\"}},\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7c9ac354",
   "metadata": {},
   "source": [
    "#### List the paths\n",
    "- $X \\rightarrow Z \\rightarrow Y$\n",
    "  - causal, leave open\n",
    "- $X \\rightarrow Z \\leftarrow u \\rightarrow Y$\n",
    "  - backdoor, but only if stratifying by Z\n",
    "- There **is no backdoor path, so no need to stratify by Z**\n",
    "- Can measure total effect of $X$ on $Y$, but not direct effect, because of Mediatior $Z$"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a96afd09",
   "metadata": {},
   "source": [
    "No backdoor path here, so no need to control for any confounds. In fact, stratifying by `Z` (the bad mediator) will introduce bias in estimate because it introduces the causal effect of `u` that would otherwise be blocked.\n",
    "\n",
    "`Z` is often a post-treatment variable, e.g. below, where \"Happiness\" is affected by the treatment \"Win Lottery\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3dceb094",
   "metadata": {},
   "source": [
    "### Verify bad mediatior with simulation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "29f99386",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"337pt\" height=\"170pt\"\n",
       " viewBox=\"0.00 0.00 337.00 169.89\" 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(4 165.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-165.89 333,-165.89 333,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"67\" cy=\"-85.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-81.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"204.5\" cy=\"-66.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"204.5\" y=\"-62.9\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Z -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M93.8,-81.83C114.43,-78.93 143.61,-74.84 166.68,-71.61\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"166.87,-75.12 176.29,-70.26 165.9,-68.18 166.87,-75.12\"/>\n",
       "<text text-anchor=\"middle\" x=\"155.75\" y=\"-78.65\" font-family=\"Times,serif\" font-size=\"14.00\">1</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"302\" cy=\"-25.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"302\" y=\"-21.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;Y -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>Z&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M228.13,-56.73C239.94,-51.66 254.61,-45.36 267.68,-39.75\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"269,-42.99 276.81,-35.83 266.24,-36.56 269,-42.99\"/>\n",
       "<text text-anchor=\"middle\" x=\"253.25\" y=\"-52.65\" font-family=\"Times,serif\" font-size=\"14.00\">1</text>\n",
       "</g>\n",
       "<!-- u -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>u</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"67\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">u</text>\n",
       "</g>\n",
       "<!-- u&#45;&gt;Z -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>u&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M92.23,-26.77C110.76,-32.49 136.88,-40.81 159.5,-48.95 162.98,-50.2 166.59,-51.55 170.19,-52.93\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"168.77,-56.13 179.36,-56.52 171.33,-49.61 168.77,-56.13\"/>\n",
       "<text text-anchor=\"middle\" x=\"155.75\" y=\"-54.65\" font-family=\"Times,serif\" font-size=\"14.00\">1</text>\n",
       "</g>\n",
       "<!-- u&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>u&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M94.35,-18.56C126.68,-17.66 183.15,-16.69 231.5,-18.95 241.96,-19.43 253.29,-20.33 263.7,-21.31\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"263.16,-24.78 273.46,-22.28 263.85,-17.81 263.16,-24.78\"/>\n",
       "<text text-anchor=\"middle\" x=\"204.5\" y=\"-24.65\" font-family=\"Times,serif\" font-size=\"14.00\">1</text>\n",
       "</g>\n",
       "<!-- unobserved -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>unobserved</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"67\" cy=\"-142.45\" rx=\"67\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-138.9\" font-family=\"Times,serif\" font-size=\"14.00\">unobserved</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0958f20>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Z\"),\n",
    "        (\"Z\", \"Y\"),\n",
    "        (\"u\", \"Z\"),\n",
    "        (\"u\", \"Y\"),\n",
    "    ],\n",
    "    node_props={\n",
    "        \"X\": {\"color\": \"red\"},\n",
    "        \"Y\": {\"color\": \"red\"},\n",
    "        \"u\": {\"style\": \"dashed\"},\n",
    "        \"unobserved\": {\"style\": \"dashed\"},\n",
    "    },\n",
    "    edge_props={\n",
    "        (\"X\", \"Z\"): {\"color\": \"red\", \"label\": \"1\"},\n",
    "        (\"Z\", \"Y\"): {\"color\": \"red\", \"label\": \"1\"},\n",
    "        (\"u\", \"Z\"): {\"label\": \"1\"},\n",
    "        (\"u\", \"Y\"): {\"label\": \"1\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "f278815d",
   "metadata": {},
   "outputs": [],
   "source": [
    "def simulate_bad_mediator(beta_XZ, beta_ZY, n_samples=100):\n",
    "    # independent variables\n",
    "    u = stats.norm.rvs(size=n_samples)\n",
    "    X = stats.norm.rvs(size=n_samples)\n",
    "\n",
    "    # causal effect of X on Z\n",
    "    mu_Z = X * beta_XZ + u\n",
    "    Z = stats.norm.rvs(size=n_samples, loc=mu_Z)\n",
    "\n",
    "    # Causal effect of Z on Y (including confound)\n",
    "    mu_Y = Z * beta_ZY + u\n",
    "    Y = stats.norm.rvs(size=n_samples, loc=mu_Y)\n",
    "\n",
    "    # Put data into format for statsmodels\n",
    "    data = pd.DataFrame(np.vstack([Y, X, Z, u]).T, columns=[\"Y\", \"X\", \"Z\", \"u\"])\n",
    "\n",
    "    unstratified_model = smf.ols(\"Y ~ X\", data=data).fit()\n",
    "    stratified_model = smf.ols(\"Y ~ X + Z\", data=data).fit()\n",
    "\n",
    "    return unstratified_model.params.X, stratified_model.params.X\n",
    "\n",
    "\n",
    "def run_bad_mediator_simulation(\n",
    "    beta_XZ=1, beta_ZY=1, n_simulations=500, n_samples_per_simualtion=100\n",
    "):\n",
    "    beta_X = beta_XZ * beta_ZY\n",
    "\n",
    "    simulations = np.array(\n",
    "        [\n",
    "            simulate_bad_mediator(\n",
    "                beta_XZ=beta_XZ, beta_ZY=beta_ZY, n_samples=n_samples_per_simualtion\n",
    "            )\n",
    "            for _ in range(n_simulations)\n",
    "        ]\n",
    "    )\n",
    "    _, ax = plt.subplots(figsize=(8, 4))\n",
    "    az.plot_dist(simulations[:, 0], label=\"Y ~ X\\ncorrect\", color=\"black\", ax=ax)\n",
    "    az.plot_dist(simulations[:, 1], label=\"Y ~ X + Z\\nwrong\", color=\"C0\", ax=ax)\n",
    "    plt.axvline(beta_X, color=\"k\", linestyle=\"--\", label=f\"actual={beta_X}\")\n",
    "    plt.legend(loc=\"upper left\")\n",
    "    plt.xlabel(\"posterior mean\")\n",
    "    plt.ylabel(\"density\");"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b051efac-af2b-4017-9861-6c22559951ad",
   "metadata": {},
   "source": [
    "#### Run the simulation, $\\beta_{XZ} = \\beta_{ZY} = 1$ "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "afdc37c2-e2ac-42ee-a2ab-f16eea466fad",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "run_bad_mediator_simulation(beta_XZ=1, beta_ZY=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "63141a0b-8e06-4641-9f66-8f8ffe6642b3",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"337pt\" height=\"170pt\"\n",
       " viewBox=\"0.00 0.00 337.00 169.89\" 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(4 165.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-165.89 333,-165.89 333,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"67\" cy=\"-85.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-81.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"204.5\" cy=\"-66.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"204.5\" y=\"-62.9\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Z -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M93.8,-81.83C114.43,-78.93 143.61,-74.84 166.68,-71.61\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"166.87,-75.12 176.29,-70.26 165.9,-68.18 166.87,-75.12\"/>\n",
       "<text text-anchor=\"middle\" x=\"155.75\" y=\"-78.65\" font-family=\"Times,serif\" font-size=\"14.00\">1</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"302\" cy=\"-25.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"302\" y=\"-21.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;Y -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>Z&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M228.13,-56.73C239.94,-51.66 254.61,-45.36 267.68,-39.75\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"269,-42.99 276.81,-35.83 266.24,-36.56 269,-42.99\"/>\n",
       "<text text-anchor=\"middle\" x=\"253.25\" y=\"-52.65\" font-family=\"Times,serif\" font-size=\"14.00\">0</text>\n",
       "</g>\n",
       "<!-- u -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>u</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"67\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">u</text>\n",
       "</g>\n",
       "<!-- u&#45;&gt;Z -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>u&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M92.23,-26.77C110.76,-32.49 136.88,-40.81 159.5,-48.95 162.98,-50.2 166.59,-51.55 170.19,-52.93\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"168.77,-56.13 179.36,-56.52 171.33,-49.61 168.77,-56.13\"/>\n",
       "<text text-anchor=\"middle\" x=\"155.75\" y=\"-54.65\" font-family=\"Times,serif\" font-size=\"14.00\">1</text>\n",
       "</g>\n",
       "<!-- u&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>u&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M94.35,-18.56C126.68,-17.66 183.15,-16.69 231.5,-18.95 241.96,-19.43 253.29,-20.33 263.7,-21.31\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"263.16,-24.78 273.46,-22.28 263.85,-17.81 263.16,-24.78\"/>\n",
       "<text text-anchor=\"middle\" x=\"204.5\" y=\"-24.65\" font-family=\"Times,serif\" font-size=\"14.00\">1</text>\n",
       "</g>\n",
       "<!-- unobserved -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>unobserved</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"67\" cy=\"-142.45\" rx=\"67\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-138.9\" font-family=\"Times,serif\" font-size=\"14.00\">unobserved</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0949b80>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Z\"),\n",
    "        (\"Z\", \"Y\"),\n",
    "        (\"u\", \"Z\"),\n",
    "        (\"u\", \"Y\"),\n",
    "    ],\n",
    "    node_props={\n",
    "        \"X\": {\"color\": \"red\"},\n",
    "        \"Y\": {\"color\": \"red\"},\n",
    "        \"u\": {\"style\": \"dashed\"},\n",
    "        \"unobserved\": {\"style\": \"dashed\"},\n",
    "    },\n",
    "    edge_props={\n",
    "        (\"X\", \"Z\"): {\"color\": \"red\", \"label\": \"1\"},\n",
    "        (\"Z\", \"Y\"): {\"color\": \"red\", \"label\": \"0\"},\n",
    "        (\"u\", \"Z\"): {\"label\": \"1\"},\n",
    "        (\"u\", \"Y\"): {\"label\": \"1\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c4cb6a05-5445-4e23-aaf9-58f1aebde3f1",
   "metadata": {},
   "source": [
    "#### Turn off Causal effect by changing $\\beta_{ZY}$ to 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "780bdbcd-df3f-4bf4-bc74-22bc704f285d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "run_bad_mediator_simulation(beta_XZ=1, beta_ZY=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9bcf031e-d74e-40e0-9c84-db7737d8df11",
   "metadata": {},
   "source": [
    "though there is no causal effect, you end up concluding a negative effect of X on Y"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bcc8d701",
   "metadata": {},
   "source": [
    "## Colliders & Descendants\n",
    "\n",
    "Generally, **Avoid the Collider**!\n",
    "\n",
    "Adding descendants of the target variable is almost always a terrible idea, because your selecting groups based on the outcome. This is known as **Case Control Bias** (selection on outcome)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "c43f34e6",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"89pt\" height=\"197pt\"\n",
       " viewBox=\"0.00 0.00 89.00 196.67\" 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(4 192.67)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-192.67 85,-192.67 85,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"54\" cy=\"-169.23\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"54\" y=\"-165.68\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"27\" cy=\"-94.34\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-90.79\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M47.33,-150.21C44.43,-142.38 40.96,-133.02 37.71,-124.26\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"41.07,-123.23 34.31,-115.07 34.5,-125.67 41.07,-123.23\"/>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"54\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"54\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Z -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>X&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M57.84,-149.67C59.82,-139.19 62.01,-125.81 63,-113.78 64.41,-96.55 64.41,-92.12 63,-74.89 62.34,-66.81 61.13,-58.12 59.81,-50.1\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"63.29,-49.73 58.11,-40.49 56.4,-50.95 63.29,-49.73\"/>\n",
       "</g>\n",
       "<!-- Y&#45;&gt;Z -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>Y&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M33.67,-75.32C36.57,-67.49 40.04,-58.13 43.29,-49.37\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"46.5,-50.77 46.69,-40.18 39.93,-48.34 46.5,-50.77\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0971e20>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[(\"X\", \"Y\"), (\"X\", \"Z\"), (\"Y\", \"Z\")],\n",
    "    node_props={\"X\": {\"color\": \"red\"}, \"Y\": {\"color\": \"red\"}},\n",
    "    edge_props={(\"X\", \"Y\"): {\"color\": \"red\"}},\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab5e1996",
   "metadata": {},
   "source": [
    "Colliders not always so obvious"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "5010ba68",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"286pt\" height=\"122pt\"\n",
       " viewBox=\"0.00 0.00 286.00 121.78\" 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(4 117.78)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-117.78 282,-117.78 282,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"27\" cy=\"-94.34\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-90.79\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"27\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M27,-74.57C27,-67.25 27,-58.69 27,-50.56\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"30.5,-50.67 27,-40.67 23.5,-50.67 30.5,-50.67\"/>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"99\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"99\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Z -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>X&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M42.27,-77.88C52.01,-68.02 64.81,-55.06 75.79,-43.94\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"78.15,-46.53 82.69,-36.96 73.17,-41.61 78.15,-46.53\"/>\n",
       "</g>\n",
       "<!-- u -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>u</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"99\" cy=\"-94.34\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"99\" y=\"-90.79\" font-family=\"Times,serif\" font-size=\"14.00\">u</text>\n",
       "</g>\n",
       "<!-- u&#45;&gt;Y -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>u&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M83.73,-77.88C73.99,-68.02 61.19,-55.06 50.21,-43.94\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"52.83,-41.61 43.31,-36.96 47.85,-46.53 52.83,-41.61\"/>\n",
       "</g>\n",
       "<!-- u&#45;&gt;Z -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>u&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M99,-74.57C99,-67.25 99,-58.69 99,-50.56\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"102.5,-50.67 99,-40.67 95.5,-50.67 102.5,-50.67\"/>\n",
       "</g>\n",
       "<!-- unobserved -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>unobserved</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"211\" cy=\"-94.34\" rx=\"67\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"211\" y=\"-90.79\" font-family=\"Times,serif\" font-size=\"14.00\">unobserved</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0a9a4b0>"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"X\", \"Z\"),\n",
    "        (\"u\", \"Y\"),\n",
    "        (\"u\", \"Z\"),\n",
    "    ],\n",
    "    node_props={\n",
    "        \"X\": {\"color\": \"red\"},\n",
    "        \"Y\": {\"color\": \"red\"},\n",
    "        \"u\": {\"style\": \"dashed\"},\n",
    "        \"unobserved\": {\"style\": \"dashed\"},\n",
    "    },\n",
    "    edge_props={(\"X\", \"Y\"): {\"color\": \"red\"}},\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6a95792a",
   "metadata": {},
   "source": [
    "Collider is formed by unobserved variable u"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "293d7935",
   "metadata": {},
   "source": [
    "### Bad Descendent: Selection on Outcome (Case Control Bias)\n",
    "\n",
    "Stratifying on a variable affected by the outcome is a **very bad** practice.\n",
    "- reduces variation in $Y$ that could have been explained by $X$\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "6b157c1a",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"486pt\" height=\"47pt\"\n",
       " viewBox=\"0.00 0.00 486.23 46.89\" 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 42.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-42.89 482.23,-42.89 482.23,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"70.18\" cy=\"-19.45\" rx=\"70.18\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"70.18\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">education, X</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"251.84\" cy=\"-19.45\" rx=\"75.48\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"251.84\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">occupation, Y</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M140.64,-19.45C148.5,-19.45 156.57,-19.45 164.6,-19.45\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"164.39,-22.95 174.39,-19.45 164.39,-15.95 164.39,-22.95\"/>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"420.78\" cy=\"-19.45\" rx=\"57.45\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"420.78\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">income, Z</text>\n",
       "</g>\n",
       "<!-- Y&#45;&gt;Z -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>Y&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M327.52,-19.45C335.57,-19.45 343.73,-19.45 351.7,-19.45\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"351.68,-22.95 361.68,-19.45 351.68,-15.95 351.68,-22.95\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0aaa720>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"Y\", \"Z\"),\n",
    "    ],\n",
    "    node_props={\n",
    "        \"X\": {\"color\": \"red\", \"label\": \"education, X\"},\n",
    "        \"Y\": {\"color\": \"red\", \"label\": \"occupation, Y\"},\n",
    "        \"Z\": {\"label\": \"income, Z\"},\n",
    "    },\n",
    "    edge_props={(\"X\", \"Y\"): {\"color\": \"red\"}},\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "03834a5d",
   "metadata": {},
   "source": [
    "#### Verify via simulation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "d724b1c7",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"257pt\" height=\"47pt\"\n",
       " viewBox=\"0.00 0.00 257.00 46.95\" 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 42.95)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-42.95 253,-42.95 253,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"27\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"124.5\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"124.5\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M54.35,-19.45C64.06,-19.45 75.25,-19.45 85.77,-19.45\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"85.72,-22.95 95.72,-19.45 85.72,-15.95 85.72,-22.95\"/>\n",
       "<text text-anchor=\"middle\" x=\"75.75\" y=\"-25.65\" font-family=\"Times,serif\" font-size=\"14.00\">1</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"222\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"222\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- Y&#45;&gt;Z -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>Y&#45;&gt;Z</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M151.85,-19.45C161.56,-19.45 172.75,-19.45 183.27,-19.45\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"183.22,-22.95 193.22,-19.45 183.22,-15.95 183.22,-22.95\"/>\n",
       "<text text-anchor=\"middle\" x=\"173.25\" y=\"-25.65\" font-family=\"Times,serif\" font-size=\"14.00\">1</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc09725a0>"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"Y\", \"Z\"),\n",
    "    ],\n",
    "    node_props={\n",
    "        \"X\": {\"color\": \"red\"},\n",
    "        \"Y\": {\"color\": \"red\"},\n",
    "    },\n",
    "    edge_props={(\"X\", \"Y\"): {\"color\": \"red\", \"label\": \"1\"}, (\"Y\", \"Z\"): {\"label\": \"1\"}},\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "8d5f1927",
   "metadata": {},
   "outputs": [],
   "source": [
    "def simulate_case_control_bias(beta_XY=1, beta_YZ=1, n_samples=100):\n",
    "    # independent variables\n",
    "    X = stats.norm.rvs(size=n_samples)\n",
    "\n",
    "    # Causal effect of Z on Y (including confound)\n",
    "    mu_Y = X * beta_XY\n",
    "    Y = stats.norm.rvs(size=n_samples, loc=mu_Y)\n",
    "\n",
    "    # causal effect of X on Z\n",
    "    mu_Z = Y * beta_YZ\n",
    "    Z = stats.norm.rvs(size=n_samples, loc=mu_Z)\n",
    "\n",
    "    # Put data into format for statsmodels\n",
    "    data = pd.DataFrame(np.vstack([Y, X, Z]).T, columns=[\"Y\", \"X\", \"Z\"])\n",
    "\n",
    "    unstratified_model = smf.ols(\"Y ~ X\", data=data).fit()\n",
    "    stratified_model = smf.ols(\"Y ~ X + Z\", data=data).fit()\n",
    "\n",
    "    return unstratified_model.params.X, stratified_model.params.X\n",
    "\n",
    "\n",
    "def run_case_control_simulation(\n",
    "    beta_XY=1, beta_YZ=1, n_simulations=500, n_samples_per_simualtion=100\n",
    "):\n",
    "    beta_X = beta_XY\n",
    "\n",
    "    simulations = np.array(\n",
    "        [\n",
    "            simulate_case_control_bias(\n",
    "                beta_XY=beta_XY, beta_YZ=beta_YZ, n_samples=n_samples_per_simualtion\n",
    "            )\n",
    "            for _ in range(n_simulations)\n",
    "        ]\n",
    "    )\n",
    "    _, ax = plt.subplots(figsize=(8, 4))\n",
    "    az.plot_dist(simulations[:, 0], label=\"Y ~ X\\ncorrect\", color=\"black\", ax=ax)\n",
    "    az.plot_dist(simulations[:, 1], label=\"Y ~ X + Z\\nwrong\", color=\"C0\", ax=ax)\n",
    "    plt.axvline(beta_X, color=\"k\", linestyle=\"--\", label=f\"actual={beta_X}\")\n",
    "    plt.legend(loc=\"upper left\")\n",
    "    plt.xlabel(\"posterior mean\")\n",
    "    plt.ylabel(\"density\");"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "55d5fe6b-6fe5-4683-8747-d00467894cb1",
   "metadata": {},
   "source": [
    "#### Descendant explains away some of he causal effect of $X$ on $Y$\n",
    "The estimated causal effect has been reduced because the descendent reduces the variation in $Y$ that can can be explained by $X$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "b2ced62d-c67a-49af-a90f-25389fb73286",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "run_case_control_simulation(beta_XY=1, beta_YZ=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "26c1859b-d03b-4050-aecd-3a80e9b44bda",
   "metadata": {},
   "source": [
    "### Removing Descendent effect by setting $\\beta_{YZ}=0$\n",
    "The descendant no longer has any effect here, so we should recover the same (correct) inference for both stratified and unstratified models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "7eedfcf6-4096-479b-b743-460a69fcf26f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "run_case_control_simulation(beta_XY=1, beta_YZ=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0f2fcae2",
   "metadata": {},
   "source": [
    "### Bad Ancestor (aka precision parasite)\n",
    "\n",
    "Now $Z$ is a parent of $X$\n",
    "\n",
    "- no backdoor path, $X$ is directly connected to $Y$\n",
    "- when stratifying by $Z$ you're explaining away variation in $X$, thus reducing the amount of causal information between $X$ and $Y$ that can be explained otherwise.\n",
    "- **Does not bias your estimate,** but it  **reduces precision**, so estimates will have more uncertainty"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "ee2dc9c6",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"257pt\" height=\"47pt\"\n",
       " viewBox=\"0.00 0.00 257.00 46.95\" 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 42.95)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-42.95 253,-42.95 253,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"124.5\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"124.5\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"222\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"222\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M151.85,-19.45C161.56,-19.45 172.75,-19.45 183.27,-19.45\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"183.22,-22.95 193.22,-19.45 183.22,-15.95 183.22,-22.95\"/>\n",
       "<text text-anchor=\"middle\" x=\"173.25\" y=\"-25.65\" font-family=\"Times,serif\" font-size=\"14.00\">1</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>Z&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M54.35,-19.45C64.06,-19.45 75.25,-19.45 85.77,-19.45\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"85.72,-22.95 95.72,-19.45 85.72,-15.95 85.72,-22.95\"/>\n",
       "<text text-anchor=\"middle\" x=\"75.75\" y=\"-25.65\" font-family=\"Times,serif\" font-size=\"14.00\">1</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0aa9940>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"Z\", \"X\"),\n",
    "    ],\n",
    "    node_props={\n",
    "        \"X\": {\"color\": \"red\"},\n",
    "        \"Y\": {\"color\": \"red\"},\n",
    "    },\n",
    "    edge_props={(\"X\", \"Y\"): {\"color\": \"red\", \"label\": \"1\"}, (\"Z\", \"X\"): {\"label\": \"1\"}},\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4fdb6677",
   "metadata": {},
   "source": [
    "#### Verify via simulation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "5eb6d0d7",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"251pt\" height=\"47pt\"\n",
       " viewBox=\"0.00 0.00 250.50 46.95\" 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 42.95)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-42.95 246.5,-42.95 246.5,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"118\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"118\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"215.5\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"215.5\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M145.35,-19.45C155.06,-19.45 166.25,-19.45 176.77,-19.45\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"176.72,-22.95 186.72,-19.45 176.72,-15.95 176.72,-22.95\"/>\n",
       "<text text-anchor=\"middle\" x=\"166.75\" y=\"-25.65\" font-family=\"Times,serif\" font-size=\"14.00\">1</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>Z&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M54.22,-19.45C62.12,-19.45 70.94,-19.45 79.43,-19.45\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"79.14,-22.95 89.14,-19.45 79.14,-15.95 79.14,-22.95\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0a8af30>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"Z\", \"X\"),\n",
    "    ],\n",
    "    node_props={\n",
    "        \"X\": {\"color\": \"red\"},\n",
    "        \"Y\": {\"color\": \"red\"},\n",
    "    },\n",
    "    edge_props={\n",
    "        (\"X\", \"Y\"): {\"color\": \"red\", \"label\": \"1\"},\n",
    "        (\"Y\", \"Z\"): {\"color\": \"red\", \"label\": \"1\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "b8abf3f8",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(123)\n",
    "\n",
    "\n",
    "def simulate_bad_ancestor(beta_ZX=1, beta_XY=1, unobserved_variance=None, n_samples=100):\n",
    "    Z = stats.norm.rvs(size=n_samples)\n",
    "\n",
    "    mu_X = Z * beta_ZX\n",
    "    if unobserved_variance is not None:\n",
    "        u = stats.norm.rvs(size=n_samples) * unobserved_variance\n",
    "        mu_X += u\n",
    "\n",
    "    X = stats.norm.rvs(size=n_samples, loc=mu_X)\n",
    "\n",
    "    mu_Y = X * beta_XY\n",
    "    if unobserved_variance is not None:\n",
    "        mu_Y += u\n",
    "\n",
    "    Y = stats.norm.rvs(size=n_samples, loc=mu_Y)\n",
    "\n",
    "    data = pd.DataFrame(np.vstack([X, Y, Z]).T, columns=[\"X\", \"Y\", \"Z\"])\n",
    "\n",
    "    non_stratified_model = smf.ols(\"Y ~ X\", data=data).fit()\n",
    "    stratified_model = smf.ols(\"Y ~ X + Z\", data=data).fit()\n",
    "\n",
    "    return non_stratified_model.params.X, stratified_model.params.X\n",
    "\n",
    "\n",
    "def run_bad_ancestor_simulation(\n",
    "    beta_ZX=1, beta_XY=1, n_simulations=500, unobserved_variance=None, n_samples_per_simualtion=100\n",
    "):\n",
    "    beta_X = beta_XY\n",
    "\n",
    "    simulations = np.array(\n",
    "        [\n",
    "            simulate_bad_ancestor(\n",
    "                beta_ZX=beta_ZX,\n",
    "                beta_XY=beta_XY,\n",
    "                unobserved_variance=unobserved_variance,\n",
    "                n_samples=n_samples_per_simualtion,\n",
    "            )\n",
    "            for _ in range(n_simulations)\n",
    "        ]\n",
    "    )\n",
    "    _, ax = plt.subplots(figsize=(8, 4))\n",
    "    az.plot_dist(simulations[:, 0], label=\"Y ~ X\\ncorrect\", color=\"black\", ax=ax)\n",
    "    az.plot_dist(simulations[:, 1], label=\"Y ~ X + Z\\nwrong\", color=\"C0\", ax=ax)\n",
    "    plt.axvline(beta_X, color=\"k\", linestyle=\"--\", label=f\"actual={beta_X}\")\n",
    "    plt.legend(loc=\"upper left\")\n",
    "    plt.xlabel(\"posterior mean\")\n",
    "    plt.ylabel(\"density\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "3f0e5ed9-53b2-45e3-995c-4183ab1d9706",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "run_bad_ancestor_simulation()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b49a25c4-cb39-4ed1-8797-f9e0fd090eda",
   "metadata": {},
   "source": [
    "Stratifying by Z doesn't add bias (it's centered on the correct value), but it does increase variance in estimator. This reduction in precision is proportional to the magnitude of the causal relationship between Z and X"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f70fc97e-4987-41de-a0ea-b34ee27eefc8",
   "metadata": {},
   "source": [
    "#### Increasing the relationshipe between Z and X further reduces precicision"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "549524a3-d2a6-453c-bc12-f0607734152d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "run_bad_ancestor_simulation(beta_ZX=3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "78814ff3",
   "metadata": {},
   "source": [
    "## Bias Amplification\n",
    "\n",
    "Stratifying on an ancestor when there are other confounders, particularly unobserved forks. **This is like the Bias Parasite scenario, but it also adds bias.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "dd123012",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"322pt\" height=\"161pt\"\n",
       " viewBox=\"0.00 0.00 322.00 160.89\" 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(4 156.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-156.89 318,-156.89 318,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"197\" cy=\"-66.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"197\" y=\"-62.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"287\" cy=\"-42.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"287\" y=\"-38.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M222.51,-59.76C231.11,-57.41 240.97,-54.73 250.3,-52.18\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"251.12,-55.58 259.85,-49.58 249.28,-48.83 251.12,-55.58\"/>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"67\" cy=\"-76.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-72.9\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>Z&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M94.21,-74.4C112.76,-72.95 137.91,-70.98 158.57,-69.37\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"158.61,-72.88 168.31,-68.61 158.07,-65.9 158.61,-72.88\"/>\n",
       "</g>\n",
       "<!-- u -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>u</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"67\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">u</text>\n",
       "</g>\n",
       "<!-- u&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>u&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M91.48,-28.07C111.23,-35.32 139.74,-45.79 162,-53.96\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"160.65,-57.2 171.25,-57.36 163.07,-50.63 160.65,-57.2\"/>\n",
       "</g>\n",
       "<!-- u&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>u&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M93.86,-22.17C132.19,-26.22 204.29,-33.82 248.44,-38.48\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"248.01,-41.96 258.32,-39.53 248.75,-35 248.01,-41.96\"/>\n",
       "</g>\n",
       "<!-- unobserved -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>unobserved</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"67\" cy=\"-133.45\" rx=\"67\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-129.9\" font-family=\"Times,serif\" font-size=\"14.00\">unobserved</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0a0eae0>"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[(\"X\", \"Y\"), (\"Z\", \"X\"), (\"u\", \"X\"), (\"u\", \"Y\")],\n",
    "    node_props={\n",
    "        \"X\": {\"color\": \"red\"},\n",
    "        \"Y\": {\"color\": \"red\"},\n",
    "        \"u\": {\"style\": \"dashed\"},\n",
    "        \"unobserved\": {\"style\": \"dashed\"},\n",
    "    },\n",
    "    edge_props={(\"X\", \"Y\"): {\"color\": \"red\"}, (\"Y\", \"Z\"): {\"color\": \"red\"}},\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0681fc5",
   "metadata": {},
   "source": [
    "### Verify via simulation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "bee895a8",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"337pt\" height=\"162pt\"\n",
       " viewBox=\"0.00 0.00 337.00 161.89\" 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(4 157.89)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-157.89 333,-157.89 333,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"204.5\" cy=\"-32.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"204.5\" y=\"-28.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"302\" cy=\"-55.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"302\" y=\"-51.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M230.59,-38.49C241.12,-41.02 253.56,-44.02 265.03,-46.78\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"264.06,-50.15 274.6,-49.09 265.7,-43.34 264.06,-50.15\"/>\n",
       "<text text-anchor=\"middle\" x=\"253.25\" y=\"-50.65\" font-family=\"Times,serif\" font-size=\"14.00\">0</text>\n",
       "</g>\n",
       "<!-- Z -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>Z</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"67\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">Z</text>\n",
       "</g>\n",
       "<!-- Z&#45;&gt;X -->\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>Z&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M94.15,-18.22C112.52,-17.77 137.62,-17.95 159.5,-20.95 162.3,-21.33 165.19,-21.83 168.07,-22.41\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"167.03,-25.76 177.56,-24.61 168.61,-18.94 167.03,-25.76\"/>\n",
       "<text text-anchor=\"middle\" x=\"155.75\" y=\"-26.65\" font-family=\"Times,serif\" font-size=\"14.00\">1</text>\n",
       "</g>\n",
       "<!-- u -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>u</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"67\" cy=\"-77.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-73.9\" font-family=\"Times,serif\" font-size=\"14.00\">u</text>\n",
       "</g>\n",
       "<!-- u&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>u&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M91.91,-69.5C113.16,-62.45 144.5,-52.04 168.55,-44.05\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"169.54,-47.41 177.93,-40.94 167.34,-40.77 169.54,-47.41\"/>\n",
       "</g>\n",
       "<!-- u&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>u&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M94.43,-77.1C131.1,-76.32 199.4,-73.76 257,-65.45 259.61,-65.07 262.3,-64.62 265.01,-64.12\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"265.42,-67.6 274.51,-62.16 264.01,-60.75 265.42,-67.6\"/>\n",
       "</g>\n",
       "<!-- unobserved -->\n",
       "<g id=\"node5\" class=\"node\">\n",
       "<title>unobserved</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" stroke-dasharray=\"5,2\" cx=\"67\" cy=\"-134.45\" rx=\"67\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-130.9\" font-family=\"Times,serif\" font-size=\"14.00\">unobserved</text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0a0eb40>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[(\"X\", \"Y\"), (\"Z\", \"X\"), (\"u\", \"X\"), (\"u\", \"Y\")],\n",
    "    node_props={\n",
    "        \"X\": {\"color\": \"red\"},\n",
    "        \"Y\": {\"color\": \"red\"},\n",
    "        \"u\": {\"style\": \"dashed\"},\n",
    "        \"unobserved\": {\"style\": \"dashed\"},\n",
    "    },\n",
    "    edge_props={\n",
    "        (\"X\", \"Y\"): {\"color\": \"red\", \"label\": \"0\"},\n",
    "        (\"Z\", \"X\"): {\"color\": \"red\", \"label\": \"1\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb4309a2-fc3c-43d9-8222-09fd62923c3c",
   "metadata": {},
   "source": [
    "#### Run simulation with no actual causal effect, $\\beta_{XY} = 0$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "90aa5209-8026-4ee2-a02d-f37e64907662",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "run_bad_ancestor_simulation(beta_ZX=1, beta_XY=0, unobserved_variance=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "02d372b2",
   "metadata": {},
   "source": [
    "Above we see that both estimators are biased -- even in the best case, we can't observe, and thus control for the the confound $u$. But when stratifying by the ancestor, things are MUCH WORSE.\n",
    "\n",
    "Hand-wavy explanation:\n",
    "- in order for $X$ and $Y$ to be associated, their causes need to be associated\n",
    "- by stratifying by $Z$, we remove the amount of variation in $X$ that is caused by $Z$\n",
    "- this reduction in variation in $X$ makes the confound $u$ more important comparatively\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "df653d2d",
   "metadata": {},
   "source": [
    "### Disrete example of bias amplificiation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "fd86c4c4",
   "metadata": {},
   "outputs": [],
   "source": [
    "def simulate_discrete_bias_amplifications(beta_ZX=1, beta_XY=1, n_samples=1000):\n",
    "    Z = stats.bernoulli.rvs(p=0.5, size=n_samples)\n",
    "    u = stats.norm.rvs(size=n_samples)\n",
    "\n",
    "    mu_X = beta_ZX * Z + u\n",
    "    X = stats.norm.rvs(loc=mu_X, size=n_samples)\n",
    "\n",
    "    mu_Y = X * beta_XY + u\n",
    "    Y = stats.norm.rvs(loc=mu_Y, size=n_samples)\n",
    "\n",
    "    data = pd.DataFrame(np.vstack([u, Z, X, Y]).T, columns=[\"u\", \"Z\", \"X\", \"Y\"])\n",
    "\n",
    "    models = {}\n",
    "    models[\"unstratified\"] = smf.ols(\"Y ~ X\", data=data).fit()\n",
    "    models[\"z=0\"] = smf.ols(\"Y ~ X\", data=data[data.Z == 0]).fit()\n",
    "    models[\"z=1\"] = smf.ols(\"Y ~ X\", data=data[data.Z == 1]).fit()\n",
    "\n",
    "    return models, data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "e4b70b72",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "beta_ZX = 7\n",
    "beta_ZY = 0\n",
    "models, data = simulate_discrete_bias_amplifications(beta_ZX=beta_ZX, beta_XY=beta_XY)\n",
    "\n",
    "fig, axs = plt.subplots(figsize=(5, 5))\n",
    "\n",
    "\n",
    "def plot_linear_model(model, color, label):\n",
    "    params = model.params\n",
    "    xs = np.linspace(data.X.min(), data.X.max(), 10)\n",
    "    ys = params.Intercept + params.X * xs\n",
    "    utils.plot_line(xs, ys, color=color, label=label)\n",
    "\n",
    "\n",
    "for z in [0, 1]:\n",
    "    color = f\"C{z}\"\n",
    "    utils.plot_scatter(data.X[data.Z == z], data.Y[data.Z == z], color=color)\n",
    "    model = models[f\"z={z}\"]\n",
    "    plot_linear_model(model, color=color, label=f\"Z={z}\")\n",
    "\n",
    "model = models[\"unstratified\"]\n",
    "plot_linear_model(model, color=\"k\", label=f\"total sample\")\n",
    "\n",
    "plt.xlim([-5, 12])\n",
    "plt.ylim([-5, 5])\n",
    "\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5b483dd",
   "metadata": {},
   "source": [
    "- When ignoring Z (the ancestor), the estimate is still somewhat biased (i.e. the black slope is not flat, as it shoudl be for $\\beta_{XY}=0$)\n",
    "- but it's not nearly as bad as the individual slopes (blue/red) when stratifying by Z. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "34ed2b55-0af1-41ca-a7fd-c121d0866005",
   "metadata": {},
   "source": [
    "# Review: Good & Bad Controls\n",
    "- **Confound**: estimator design or sample that \"confoudnds\" or \"confuses\" our causal estiamte\n",
    "- **Control**: variable added to the analysis to that a causal estimate is possible\n",
    "- Adding controls can often be worse than ommitting them\n",
    "- Make assumptions explicit, and use backdoor criterion to verify those assumptions\n",
    "\n",
    "**You have to do scientific modeling to do scientific analysis**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "665a9678",
   "metadata": {},
   "source": [
    "# BONUS: The Table 2 Fallacy\n",
    "- Not all coefficients represent causal effects, particularly those in the adjustment set\n",
    "- Those that _are_ causal effects tend to be partial effects, not total causal effects.\n",
    "- Table 2 actively _encourage_ misinterpretation\n",
    "- As mentioned multiple times: **Need different estimators for addressing different causal effects.**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9457007c",
   "metadata": {},
   "source": [
    "## Example: Smoking, Age, HIV, and Stroke"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "7794ff9c",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"513pt\" height=\"104pt\"\n",
       " viewBox=\"0.00 0.00 513.04 103.76\" 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 99.76)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-99.76 509.04,-99.76 509.04,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"322.14\" cy=\"-61.45\" rx=\"41.54\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"322.14\" y=\"-57.9\" font-family=\"Times,serif\" font-size=\"14.00\">X, HIV</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"452.36\" cy=\"-61.45\" rx=\"52.68\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"452.36\" y=\"-57.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y, Stroke</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M364.06,-61.45C371.73,-61.45 379.92,-61.45 388.08,-61.45\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"388.05,-64.95 398.05,-61.45 388.05,-57.95 388.05,-64.95\"/>\n",
       "</g>\n",
       "<!-- S -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>S</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"181.31\" cy=\"-19.45\" rx=\"63.29\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"181.31\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">S, Smoking</text>\n",
       "</g>\n",
       "<!-- S&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>S&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M227.38,-33.08C242.8,-37.75 260.05,-42.97 275.52,-47.64\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"274.37,-50.95 284.96,-50.5 276.4,-44.25 274.37,-50.95\"/>\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=\"M244.91,-20.7C280.08,-22.3 324.63,-25.81 363.68,-33.45 376.78,-36.01 390.67,-39.88 403.43,-43.93\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"402.28,-47.24 412.87,-47.04 404.47,-40.59 402.28,-47.24\"/>\n",
       "</g>\n",
       "<!-- A -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>A</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"41.01\" cy=\"-66.45\" rx=\"41.01\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"41.01\" y=\"-62.9\" font-family=\"Times,serif\" font-size=\"14.00\">A, Age</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;X -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>A&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M82.3,-65.72C131.53,-64.84 214.93,-63.35 269,-62.38\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"268.96,-65.88 278.89,-62.2 268.83,-58.88 268.96,-65.88\"/>\n",
       "</g>\n",
       "<!-- A&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>A&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M78.9,-74.45C138.93,-86.28 261.34,-105.39 363.68,-89.45 377.25,-87.33 391.59,-83.48 404.65,-79.3\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"405.48,-82.71 413.85,-76.22 403.25,-76.08 405.48,-82.71\"/>\n",
       "</g>\n",
       "<!-- A&#45;&gt;S -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>A&#45;&gt;S</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M75.14,-55.19C90.69,-49.9 109.61,-43.47 127.1,-37.53\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"127.98,-40.93 136.32,-34.39 125.73,-34.3 127.98,-40.93\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0b23740>"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"S\", \"Y\"),\n",
    "        (\"S\", \"X\"),\n",
    "        (\"A\", \"Y\"),\n",
    "        (\"A\", \"X\"),\n",
    "        (\"A\", \"S\"),\n",
    "    ],\n",
    "    node_props={\n",
    "        \"X\": {\"label\": \"X, HIV\", \"color\": \"red\"},\n",
    "        \"Y\": {\"label\": \"Y, Stroke\", \"color\": \"red\"},\n",
    "        \"S\": {\"label\": \"S, Smoking\"},\n",
    "        \"A\": {\"label\": \"A, Age\"},\n",
    "    },\n",
    "    edge_props={(\"X\", \"Y\"): {\"color\": \"red\"}},\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c1517e58",
   "metadata": {},
   "source": [
    "### Identify paths via backdoor criterion\n",
    "- $X \\rightarrow Y$ (front door)\n",
    "- $X \\leftarrow S \\rightarrow Y$ (backdoor, fork)\n",
    "- $X \\leftarrow A \\rightarrow Y$ (backdoor, fork)\n",
    "- $X \\leftarrow A \\rightarrow S \\rightarrow Y$ (backdoor, fork and pipe)\n",
    "\n",
    "Adjustment set is $\\{S, A\\}$\n",
    "\n",
    "### Conditional Statistical Model\n",
    "\n",
    "$$\n",
    "\\begin{align*}\n",
    "Y_i &\\sim \\mathcal{N}(\\mu_i, \\sigma) \\\\\n",
    "\\mu_i = &\\alpha + \\beta_X X_i + \\beta_S S_i + \\beta_A A_i\n",
    "\\end{align*}\n",
    "$$"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "89ac3a08",
   "metadata": {},
   "source": [
    "### Looking at the model \"from the perspective of various variables\"\n",
    "#### From perspective of X\n",
    "Conditioning on $A$ and $S$ essentially removes the arrows going into $X$, $\\beta_X$ giving us the **direct effect** of $X$ on $Y$\n",
    "  - we've removed all backdoor paths by stratifying by $S$ and $A$ throught the coefficients $\\beta_S, \\beta_A$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "b6e7d1d9",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"332pt\" height=\"103pt\"\n",
       " viewBox=\"0.00 0.00 332.00 102.61\" 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 98.61)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-98.61 328,-98.61 328,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"207\" cy=\"-61.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"207\" y=\"-57.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"297\" cy=\"-61.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"297\" y=\"-57.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M234.4,-61.45C241.89,-61.45 250.18,-61.45 258.2,-61.45\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"258.1,-64.95 268.1,-61.45 258.1,-57.95 258.1,-64.95\"/>\n",
       "</g>\n",
       "<!-- S -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>S</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"117\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"117\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">S</text>\n",
       "</g>\n",
       "<!-- S&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>S&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"none\" d=\"M140.21,-30.04C150.32,-34.87 162.48,-40.67 173.58,-45.97\"/>\n",
       "<polygon fill=\"none\" stroke=\"none\" points=\"171.9,-49.05 182.44,-50.2 174.92,-42.73 171.9,-49.05\"/>\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=\"M144.24,-20.34C168.04,-21.69 203.91,-25.05 234,-33.45 244.25,-36.31 254.96,-40.62 264.55,-45.01\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"262.9,-48.11 273.43,-49.27 265.93,-41.79 262.9,-48.11\"/>\n",
       "</g>\n",
       "<!-- A -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>A</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-66.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-62.9\" font-family=\"Times,serif\" font-size=\"14.00\">A</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;X -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>A&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"none\" d=\"M54.26,-65.71C84.37,-64.86 133.96,-63.47 168.21,-62.51\"/>\n",
       "<polygon fill=\"none\" stroke=\"none\" points=\"168.22,-66.01 178.11,-62.23 168.02,-59.01 168.22,-66.01\"/>\n",
       "</g>\n",
       "<!-- A&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>A&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M52.04,-74.1C90.74,-85.31 169.09,-103.42 234,-89.45 244.49,-87.19 255.34,-83.01 264.97,-78.55\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"266.46,-81.72 273.89,-74.16 263.37,-75.44 266.46,-81.72\"/>\n",
       "</g>\n",
       "<!-- A&#45;&gt;S -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>A&#45;&gt;S</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M49.3,-55.07C59.92,-49.39 72.98,-42.42 84.71,-36.16\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"86.08,-39.39 93.25,-31.59 82.78,-33.22 86.08,-39.39\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0c9cf80>"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"S\", \"Y\"),\n",
    "        (\"S\", \"X\"),\n",
    "        (\"A\", \"Y\"),\n",
    "        (\"A\", \"X\"),\n",
    "        (\"A\", \"S\"),\n",
    "    ],\n",
    "    node_props={\"X\": {\"color\": \"red\"}, \"Y\": {\"color\": \"red\"}},\n",
    "    edge_props={\n",
    "        (\"X\", \"Y\"): {\"color\": \"red\"},\n",
    "        (\"A\", \"X\"): {\"color\": \"none\"},\n",
    "        (\"S\", \"X\"): {\"color\": \"none\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1bc86707",
   "metadata": {},
   "source": [
    "#### From perspective of S"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "7297f8cb",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"332pt\" height=\"103pt\"\n",
       " viewBox=\"0.00 0.00 332.00 102.61\" 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 98.61)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-98.61 328,-98.61 328,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"207\" cy=\"-61.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"207\" y=\"-57.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"297\" cy=\"-61.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"297\" y=\"-57.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M234.4,-61.45C241.89,-61.45 250.18,-61.45 258.2,-61.45\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"258.1,-64.95 268.1,-61.45 258.1,-57.95 258.1,-64.95\"/>\n",
       "</g>\n",
       "<!-- S -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>S</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"117\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"117\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">S</text>\n",
       "</g>\n",
       "<!-- S&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>S&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M140.21,-30.04C150.32,-34.87 162.48,-40.67 173.58,-45.97\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"171.9,-49.05 182.44,-50.2 174.92,-42.73 171.9,-49.05\"/>\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=\"red\" d=\"M144.24,-20.34C168.04,-21.69 203.91,-25.05 234,-33.45 244.25,-36.31 254.96,-40.62 264.55,-45.01\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"262.9,-48.11 273.43,-49.27 265.93,-41.79 262.9,-48.11\"/>\n",
       "</g>\n",
       "<!-- A -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>A</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-66.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-62.9\" font-family=\"Times,serif\" font-size=\"14.00\">A</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;X -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>A&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M54.26,-65.71C84.37,-64.86 133.96,-63.47 168.21,-62.51\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"168.22,-66.01 178.11,-62.23 168.02,-59.01 168.22,-66.01\"/>\n",
       "</g>\n",
       "<!-- A&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>A&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M52.04,-74.1C90.74,-85.31 169.09,-103.42 234,-89.45 244.49,-87.19 255.34,-83.01 264.97,-78.55\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"266.46,-81.72 273.89,-74.16 263.37,-75.44 266.46,-81.72\"/>\n",
       "</g>\n",
       "<!-- A&#45;&gt;S -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>A&#45;&gt;S</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M49.3,-55.07C59.92,-49.39 72.98,-42.42 84.71,-36.16\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"86.08,-39.39 93.25,-31.59 82.78,-33.22 86.08,-39.39\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0cc24b0>"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"S\", \"Y\"),\n",
    "        (\"S\", \"X\"),\n",
    "        (\"A\", \"Y\"),\n",
    "        (\"A\", \"X\"),\n",
    "        (\"A\", \"S\"),\n",
    "    ],\n",
    "    node_props={\"Y\": {\"color\": \"red\"}, \"S\": {\"color\": \"red\"}},\n",
    "    edge_props={\n",
    "        (\"X\", \"Y\"): {\"color\": \"red\"},\n",
    "        (\"S\", \"X\"): {\"color\": \"red\"},\n",
    "        (\"S\", \"Y\"): {\"color\": \"red\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc0281f0",
   "metadata": {},
   "source": [
    "##### Adjusted graph with the full model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "5822bf10",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"332pt\" height=\"103pt\"\n",
       " viewBox=\"0.00 0.00 332.00 102.61\" 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 98.61)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-98.61 328,-98.61 328,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"207\" cy=\"-61.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"207\" y=\"-57.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"297\" cy=\"-61.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"297\" y=\"-57.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M234.4,-61.45C241.89,-61.45 250.18,-61.45 258.2,-61.45\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"258.1,-64.95 268.1,-61.45 258.1,-57.95 258.1,-64.95\"/>\n",
       "</g>\n",
       "<!-- S -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>S</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"117\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"117\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">S</text>\n",
       "</g>\n",
       "<!-- S&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>S&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"none\" d=\"M140.21,-30.04C150.32,-34.87 162.48,-40.67 173.58,-45.97\"/>\n",
       "<polygon fill=\"none\" stroke=\"none\" points=\"171.9,-49.05 182.44,-50.2 174.92,-42.73 171.9,-49.05\"/>\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=\"red\" d=\"M144.24,-20.34C168.04,-21.69 203.91,-25.05 234,-33.45 244.25,-36.31 254.96,-40.62 264.55,-45.01\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"262.9,-48.11 273.43,-49.27 265.93,-41.79 262.9,-48.11\"/>\n",
       "</g>\n",
       "<!-- A -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>A</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"27\" cy=\"-66.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-62.9\" font-family=\"Times,serif\" font-size=\"14.00\">A</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;X -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>A&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M54.26,-65.71C84.37,-64.86 133.96,-63.47 168.21,-62.51\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"168.22,-66.01 178.11,-62.23 168.02,-59.01 168.22,-66.01\"/>\n",
       "</g>\n",
       "<!-- A&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>A&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M52.04,-74.1C90.74,-85.31 169.09,-103.42 234,-89.45 244.49,-87.19 255.34,-83.01 264.97,-78.55\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"266.46,-81.72 273.89,-74.16 263.37,-75.44 266.46,-81.72\"/>\n",
       "</g>\n",
       "<!-- A&#45;&gt;S -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>A&#45;&gt;S</title>\n",
       "<path fill=\"none\" stroke=\"none\" d=\"M49.3,-55.07C59.92,-49.39 72.98,-42.42 84.71,-36.16\"/>\n",
       "<polygon fill=\"none\" stroke=\"none\" points=\"86.08,-39.39 93.25,-31.59 82.78,-33.22 86.08,-39.39\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0cb73b0>"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"S\", \"Y\"),\n",
    "        (\"S\", \"X\"),\n",
    "        (\"A\", \"Y\"),\n",
    "        (\"A\", \"X\"),\n",
    "        (\"A\", \"S\"),\n",
    "    ],\n",
    "    node_props={\"Y\": {\"color\": \"red\"}, \"S\": {\"color\": \"red\"}},\n",
    "    edge_props={\n",
    "        (\"S\", \"X\"): {\"color\": \"none\"},\n",
    "        (\"A\", \"S\"): {\"color\": \"none\"},\n",
    "        (\"S\", \"Y\"): {\"color\": \"red\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71fe3176-af3b-4a0f-a0f5-255d0f2d9512",
   "metadata": {},
   "source": [
    "- In the unconditional model, the effect of $S$ on $Y$ is confounded by $A$, because it's a common cause of $S$ and $Y$ (and $X$)\n",
    "- Conditioning on $A$ & $X$ (via the same statistical model above), $\\beta_{S}$ gives the **direct effect** of $S$ on $Y$/\n",
    "  - Since we've blocked the path along $X$ in the linear regression, we no longer get the **total effect**."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dbb087cb",
   "metadata": {},
   "source": [
    "#### From the perspective of A\n",
    "\n",
    "In the unconditional model, the **total causal effect** of $A$ on $Y$ flows through all paths:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "4d867991",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"332pt\" height=\"103pt\"\n",
       " viewBox=\"0.00 0.00 332.00 102.61\" 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 98.61)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-98.61 328,-98.61 328,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"207\" cy=\"-61.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"207\" y=\"-57.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"297\" cy=\"-61.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"297\" y=\"-57.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M234.4,-61.45C241.89,-61.45 250.18,-61.45 258.2,-61.45\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"258.1,-64.95 268.1,-61.45 258.1,-57.95 258.1,-64.95\"/>\n",
       "</g>\n",
       "<!-- S -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>S</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"117\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"117\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">S</text>\n",
       "</g>\n",
       "<!-- S&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>S&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M140.21,-30.04C150.32,-34.87 162.48,-40.67 173.58,-45.97\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"171.9,-49.05 182.44,-50.2 174.92,-42.73 171.9,-49.05\"/>\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=\"red\" d=\"M144.24,-20.34C168.04,-21.69 203.91,-25.05 234,-33.45 244.25,-36.31 254.96,-40.62 264.55,-45.01\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"262.9,-48.11 273.43,-49.27 265.93,-41.79 262.9,-48.11\"/>\n",
       "</g>\n",
       "<!-- A -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>A</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"27\" cy=\"-66.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-62.9\" font-family=\"Times,serif\" font-size=\"14.00\">A</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;X -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>A&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M54.26,-65.71C84.37,-64.86 133.96,-63.47 168.21,-62.51\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"168.22,-66.01 178.11,-62.23 168.02,-59.01 168.22,-66.01\"/>\n",
       "</g>\n",
       "<!-- A&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>A&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M52.04,-74.1C90.74,-85.31 169.09,-103.42 234,-89.45 244.49,-87.19 255.34,-83.01 264.97,-78.55\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"266.46,-81.72 273.89,-74.16 263.37,-75.44 266.46,-81.72\"/>\n",
       "</g>\n",
       "<!-- A&#45;&gt;S -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>A&#45;&gt;S</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M49.3,-55.07C59.92,-49.39 72.98,-42.42 84.71,-36.16\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"86.08,-39.39 93.25,-31.59 82.78,-33.22 86.08,-39.39\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0aefec0>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"S\", \"Y\"),\n",
    "        (\"S\", \"X\"),\n",
    "        (\"A\", \"Y\"),\n",
    "        (\"A\", \"X\"),\n",
    "        (\"A\", \"S\"),\n",
    "    ],\n",
    "    node_props={\"Y\": {\"color\": \"red\"}, \"A\": {\"color\": \"red\"}},\n",
    "    edge_props={\n",
    "        (\"X\", \"Y\"): {\"color\": \"red\"},\n",
    "        (\"A\", \"X\"): {\"color\": \"red\"},\n",
    "        (\"A\", \"Y\"): {\"color\": \"red\"},\n",
    "        (\"A\", \"S\"): {\"color\": \"red\"},\n",
    "        (\"S\", \"X\"): {\"color\": \"red\"},\n",
    "        (\"S\", \"Y\"): {\"color\": \"red\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1de302fd",
   "metadata": {},
   "source": [
    "- Conditioning on $S$ & $X$ (via the same statistical model above), $\\beta_{S}$ gives the **direct effect** of $S$ on $Y$/\n",
    "  - Since we've blocked the path along $X$ in the linear regression, we no longer get the **total effect**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "ad7c4e98",
   "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 12.0.0 (20240803.0821)\n",
       " -->\n",
       "<!-- Pages: 1 -->\n",
       "<svg width=\"332pt\" height=\"103pt\"\n",
       " viewBox=\"0.00 0.00 332.00 102.61\" 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 98.61)\">\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-98.61 328,-98.61 328,4 -4,4\"/>\n",
       "<!-- X -->\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>X</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"207\" cy=\"-61.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"207\" y=\"-57.9\" font-family=\"Times,serif\" font-size=\"14.00\">X</text>\n",
       "</g>\n",
       "<!-- Y -->\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>Y</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"297\" cy=\"-61.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"297\" y=\"-57.9\" font-family=\"Times,serif\" font-size=\"14.00\">Y</text>\n",
       "</g>\n",
       "<!-- X&#45;&gt;Y -->\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>X&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M234.4,-61.45C241.89,-61.45 250.18,-61.45 258.2,-61.45\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"258.1,-64.95 268.1,-61.45 258.1,-57.95 258.1,-64.95\"/>\n",
       "</g>\n",
       "<!-- S -->\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>S</title>\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"117\" cy=\"-19.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"117\" y=\"-15.9\" font-family=\"Times,serif\" font-size=\"14.00\">S</text>\n",
       "</g>\n",
       "<!-- S&#45;&gt;X -->\n",
       "<g id=\"edge3\" class=\"edge\">\n",
       "<title>S&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"black\" d=\"M140.21,-30.04C150.32,-34.87 162.48,-40.67 173.58,-45.97\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"171.9,-49.05 182.44,-50.2 174.92,-42.73 171.9,-49.05\"/>\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=\"M144.24,-20.34C168.04,-21.69 203.91,-25.05 234,-33.45 244.25,-36.31 254.96,-40.62 264.55,-45.01\"/>\n",
       "<polygon fill=\"black\" stroke=\"black\" points=\"262.9,-48.11 273.43,-49.27 265.93,-41.79 262.9,-48.11\"/>\n",
       "</g>\n",
       "<!-- A -->\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>A</title>\n",
       "<ellipse fill=\"none\" stroke=\"red\" cx=\"27\" cy=\"-66.45\" rx=\"27\" ry=\"19.45\"/>\n",
       "<text text-anchor=\"middle\" x=\"27\" y=\"-62.9\" font-family=\"Times,serif\" font-size=\"14.00\">A</text>\n",
       "</g>\n",
       "<!-- A&#45;&gt;X -->\n",
       "<g id=\"edge5\" class=\"edge\">\n",
       "<title>A&#45;&gt;X</title>\n",
       "<path fill=\"none\" stroke=\"none\" d=\"M54.26,-65.71C84.37,-64.86 133.96,-63.47 168.21,-62.51\"/>\n",
       "<polygon fill=\"none\" stroke=\"none\" points=\"168.22,-66.01 178.11,-62.23 168.02,-59.01 168.22,-66.01\"/>\n",
       "</g>\n",
       "<!-- A&#45;&gt;Y -->\n",
       "<g id=\"edge4\" class=\"edge\">\n",
       "<title>A&#45;&gt;Y</title>\n",
       "<path fill=\"none\" stroke=\"red\" d=\"M52.04,-74.1C90.74,-85.31 169.09,-103.42 234,-89.45 244.49,-87.19 255.34,-83.01 264.97,-78.55\"/>\n",
       "<polygon fill=\"red\" stroke=\"red\" points=\"266.46,-81.72 273.89,-74.16 263.37,-75.44 266.46,-81.72\"/>\n",
       "</g>\n",
       "<!-- A&#45;&gt;S -->\n",
       "<g id=\"edge6\" class=\"edge\">\n",
       "<title>A&#45;&gt;S</title>\n",
       "<path fill=\"none\" stroke=\"none\" d=\"M49.3,-55.07C59.92,-49.39 72.98,-42.42 84.71,-36.16\"/>\n",
       "<polygon fill=\"none\" stroke=\"none\" points=\"86.08,-39.39 93.25,-31.59 82.78,-33.22 86.08,-39.39\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>\n"
      ],
      "text/plain": [
       "<graphviz.graphs.Digraph at 0x7f5bc0aeec00>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.draw_causal_graph(\n",
    "    edge_list=[\n",
    "        (\"X\", \"Y\"),\n",
    "        (\"S\", \"Y\"),\n",
    "        (\"S\", \"X\"),\n",
    "        (\"A\", \"Y\"),\n",
    "        (\"A\", \"X\"),\n",
    "        (\"A\", \"S\"),\n",
    "    ],\n",
    "    node_props={\"Y\": {\"color\": \"red\"}, \"A\": {\"color\": \"red\"}},\n",
    "    edge_props={\n",
    "        (\"A\", \"X\"): {\"color\": \"none\"},\n",
    "        (\"A\", \"S\"): {\"color\": \"none\"},\n",
    "        (\"A\", \"Y\"): {\"color\": \"red\"},\n",
    "    },\n",
    "    graph_direction=\"LR\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "46eb240d",
   "metadata": {},
   "source": [
    "This gets trickier if we consider unobserved confounds on variables!\n",
    "\n",
    "## Summary: Table 2 Fallacy\n",
    "- Not all coefficients have the same interpretation\n",
    "  - different estimands require different models\n",
    "- Do not present coefficients as they are equal (i.e. in Table 2)\n",
    "- ...or, don't present coefficients at all, instead push out predictions.\n",
    "- Provide _explicit interpretation_ of each"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "84c9e68e",
   "metadata": {},
   "source": [
    "## Authors\n",
    "* Ported to PyMC by Dustin Stansbury (2024)\n",
    "* Based on Statistical Rethinking (2023) lectures by Richard McElreath"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "86b4b6ff",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Last updated: Tue Dec 17 2024\n",
      "\n",
      "Python implementation: CPython\n",
      "Python version       : 3.12.5\n",
      "IPython version      : 8.27.0\n",
      "\n",
      "pytensor: 2.26.4\n",
      "aeppl   : not installed\n",
      "xarray  : 2024.7.0\n",
      "\n",
      "pymc       : 5.19.1\n",
      "scipy      : 1.14.1\n",
      "statsmodels: 0.14.2\n",
      "xarray     : 2024.7.0\n",
      "matplotlib : 3.9.2\n",
      "pandas     : 2.2.2\n",
      "numpy      : 1.26.4\n",
      "arviz      : 0.19.0\n",
      "\n",
      "Watermark: 2.5.0\n",
      "\n"
     ]
    }
   ],
   "source": [
    "%load_ext watermark\n",
    "%watermark -n -u -v -iv -w -p pytensor,aeppl,xarray"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "78c80f20",
   "metadata": {},
   "source": [
    ":::{include} ../page_footer.md\n",
    ":::"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}