04-Ensemble-learning-Experiment.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "26d1037b-3887-4842-a0c8-eebb9e8da867",
   "metadata": {},
   "source": [
    "### Preamble\n",
    "\n",
    "> _If you've reached this part of the tutorial, you're a **Star**_ 🌟 🙌\n",
    "> \n",
    "> **Thank you!** I hope you're finding this useful, and interesting!\n",
    "> \n",
    "> _Valerio_"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ac1398ce",
   "metadata": {},
   "source": [
    "# Ensemble learning to combine multiple ML models\n",
    "\n",
    "In the last notebook of our tutorial, we will explore how to use [Ensemble learning](https://en.wikipedia.org/wiki/Ensemble_learning) as a potential strategy to combine multiple machine learning models, trained on different datasets."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "53e6b02a-c638-4041-a82b-b64c90066ef6",
   "metadata": {},
   "source": [
    "## Step 1. Login to datasites as **External Researcher**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8fa52e68-66f5-4b00-b724-15e3610765fd",
   "metadata": {},
   "source": [
    "Launch Datasites if not running:\n",
    "```bash\n",
    "$ python launch_datasites.py\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "ad4acae5-f788-447f-b366-0c6e15f11691",
   "metadata": {},
   "outputs": [],
   "source": [
    "import syft as sy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "ccf7693a-195a-46b8-9e44-a7fc5c773279",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logged into <Cleveland Clinic: High-side Datasite> as GUEST\n",
      "Logged into <Hungarian Inst. of Cardiology: High-side Datasite> as GUEST\n",
      "Logged into <Univ. Hospitals Zurich and Basel: High-side Datasite> as GUEST\n",
      "Logged into <V.A. Medical Center: High-side Datasite> as GUEST\n"
     ]
    }
   ],
   "source": [
    "from datasites import DATASITE_URLS\n",
    "\n",
    "datasites = {}\n",
    "for name, url in DATASITE_URLS.items():\n",
    "    datasites[name] = sy.login_as_guest(url=url)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e0d36de7-494f-401b-8f95-444cb5375809",
   "metadata": {},
   "source": [
    "## Step 2. Prepare Ensemble Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "3a725f05-8e3a-4c8c-81fc-80046bcd5936",
   "metadata": {},
   "outputs": [],
   "source": [
    "mock_data = datasites[\"Cleveland Clinic\"].datasets[\"Heart Disease Dataset\"].assets[\"Heart Study Data\"].mock"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "9d5b9ad8-8ba0-4fbf-9c21-20bf90a1bf48",
   "metadata": {},
   "outputs": [],
   "source": [
    "# load pre-trained models\n",
    "from utils import load_models\n",
    "\n",
    "models = load_models(datasites, root=\"./models\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "4a1bae21-f681-474e-8832-7ab12b0baf96",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {\n",
       "  /* Definition of color scheme common for light and dark mode */\n",
       "  --sklearn-color-text: black;\n",
       "  --sklearn-color-line: gray;\n",
       "  /* Definition of color scheme for unfitted estimators */\n",
       "  --sklearn-color-unfitted-level-0: #fff5e6;\n",
       "  --sklearn-color-unfitted-level-1: #f6e4d2;\n",
       "  --sklearn-color-unfitted-level-2: #ffe0b3;\n",
       "  --sklearn-color-unfitted-level-3: chocolate;\n",
       "  /* Definition of color scheme for fitted estimators */\n",
       "  --sklearn-color-fitted-level-0: #f0f8ff;\n",
       "  --sklearn-color-fitted-level-1: #d4ebff;\n",
       "  --sklearn-color-fitted-level-2: #b3dbfd;\n",
       "  --sklearn-color-fitted-level-3: cornflowerblue;\n",
       "\n",
       "  /* Specific color for light theme */\n",
       "  --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
       "  --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, white)));\n",
       "  --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
       "  --sklearn-color-icon: #696969;\n",
       "\n",
       "  @media (prefers-color-scheme: dark) {\n",
       "    /* Redefinition of color scheme for dark theme */\n",
       "    --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
       "    --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, #111)));\n",
       "    --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
       "    --sklearn-color-icon: #878787;\n",
       "  }\n",
       "}\n",
       "\n",
       "#sk-container-id-1 {\n",
       "  color: var(--sklearn-color-text);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 pre {\n",
       "  padding: 0;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 input.sk-hidden--visually {\n",
       "  border: 0;\n",
       "  clip: rect(1px 1px 1px 1px);\n",
       "  clip: rect(1px, 1px, 1px, 1px);\n",
       "  height: 1px;\n",
       "  margin: -1px;\n",
       "  overflow: hidden;\n",
       "  padding: 0;\n",
       "  position: absolute;\n",
       "  width: 1px;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-dashed-wrapped {\n",
       "  border: 1px dashed var(--sklearn-color-line);\n",
       "  margin: 0 0.4em 0.5em 0.4em;\n",
       "  box-sizing: border-box;\n",
       "  padding-bottom: 0.4em;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-container {\n",
       "  /* jupyter's `normalize.less` sets `[hidden] { display: none; }`\n",
       "     but bootstrap.min.css set `[hidden] { display: none !important; }`\n",
       "     so we also need the `!important` here to be able to override the\n",
       "     default hidden behavior on the sphinx rendered scikit-learn.org.\n",
       "     See: https://github.com/scikit-learn/scikit-learn/issues/21755 */\n",
       "  display: inline-block !important;\n",
       "  position: relative;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-text-repr-fallback {\n",
       "  display: none;\n",
       "}\n",
       "\n",
       "div.sk-parallel-item,\n",
       "div.sk-serial,\n",
       "div.sk-item {\n",
       "  /* draw centered vertical line to link estimators */\n",
       "  background-image: linear-gradient(var(--sklearn-color-text-on-default-background), var(--sklearn-color-text-on-default-background));\n",
       "  background-size: 2px 100%;\n",
       "  background-repeat: no-repeat;\n",
       "  background-position: center center;\n",
       "}\n",
       "\n",
       "/* Parallel-specific style estimator block */\n",
       "\n",
       "#sk-container-id-1 div.sk-parallel-item::after {\n",
       "  content: \"\";\n",
       "  width: 100%;\n",
       "  border-bottom: 2px solid var(--sklearn-color-text-on-default-background);\n",
       "  flex-grow: 1;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-parallel {\n",
       "  display: flex;\n",
       "  align-items: stretch;\n",
       "  justify-content: center;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  position: relative;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-parallel-item {\n",
       "  display: flex;\n",
       "  flex-direction: column;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-parallel-item:first-child::after {\n",
       "  align-self: flex-end;\n",
       "  width: 50%;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-parallel-item:last-child::after {\n",
       "  align-self: flex-start;\n",
       "  width: 50%;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-parallel-item:only-child::after {\n",
       "  width: 0;\n",
       "}\n",
       "\n",
       "/* Serial-specific style estimator block */\n",
       "\n",
       "#sk-container-id-1 div.sk-serial {\n",
       "  display: flex;\n",
       "  flex-direction: column;\n",
       "  align-items: center;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  padding-right: 1em;\n",
       "  padding-left: 1em;\n",
       "}\n",
       "\n",
       "\n",
       "/* Toggleable style: style used for estimator/Pipeline/ColumnTransformer box that is\n",
       "clickable and can be expanded/collapsed.\n",
       "- Pipeline and ColumnTransformer use this feature and define the default style\n",
       "- Estimators will overwrite some part of the style using the `sk-estimator` class\n",
       "*/\n",
       "\n",
       "/* Pipeline and ColumnTransformer style (default) */\n",
       "\n",
       "#sk-container-id-1 div.sk-toggleable {\n",
       "  /* Default theme specific background. It is overwritten whether we have a\n",
       "  specific estimator or a Pipeline/ColumnTransformer */\n",
       "  background-color: var(--sklearn-color-background);\n",
       "}\n",
       "\n",
       "/* Toggleable label */\n",
       "#sk-container-id-1 label.sk-toggleable__label {\n",
       "  cursor: pointer;\n",
       "  display: block;\n",
       "  width: 100%;\n",
       "  margin-bottom: 0;\n",
       "  padding: 0.5em;\n",
       "  box-sizing: border-box;\n",
       "  text-align: center;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 label.sk-toggleable__label-arrow:before {\n",
       "  /* Arrow on the left of the label */\n",
       "  content: \"▸\";\n",
       "  float: left;\n",
       "  margin-right: 0.25em;\n",
       "  color: var(--sklearn-color-icon);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {\n",
       "  color: var(--sklearn-color-text);\n",
       "}\n",
       "\n",
       "/* Toggleable content - dropdown */\n",
       "\n",
       "#sk-container-id-1 div.sk-toggleable__content {\n",
       "  max-height: 0;\n",
       "  max-width: 0;\n",
       "  overflow: hidden;\n",
       "  text-align: left;\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-toggleable__content.fitted {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-toggleable__content pre {\n",
       "  margin: 0.2em;\n",
       "  border-radius: 0.25em;\n",
       "  color: var(--sklearn-color-text);\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-toggleable__content.fitted pre {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {\n",
       "  /* Expand drop-down */\n",
       "  max-height: 200px;\n",
       "  max-width: 100%;\n",
       "  overflow: auto;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {\n",
       "  content: \"▾\";\n",
       "}\n",
       "\n",
       "/* Pipeline/ColumnTransformer-specific style */\n",
       "\n",
       "#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-label.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Estimator-specific style */\n",
       "\n",
       "/* Colorize estimator box */\n",
       "#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-estimator.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-label label.sk-toggleable__label,\n",
       "#sk-container-id-1 div.sk-label label {\n",
       "  /* The background is the default theme color */\n",
       "  color: var(--sklearn-color-text-on-default-background);\n",
       "}\n",
       "\n",
       "/* On hover, darken the color of the background */\n",
       "#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "/* Label box, darken color on hover, fitted */\n",
       "#sk-container-id-1 div.sk-label.fitted:hover label.sk-toggleable__label.fitted {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Estimator label */\n",
       "\n",
       "#sk-container-id-1 div.sk-label label {\n",
       "  font-family: monospace;\n",
       "  font-weight: bold;\n",
       "  display: inline-block;\n",
       "  line-height: 1.2em;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-label-container {\n",
       "  text-align: center;\n",
       "}\n",
       "\n",
       "/* Estimator-specific */\n",
       "#sk-container-id-1 div.sk-estimator {\n",
       "  font-family: monospace;\n",
       "  border: 1px dotted var(--sklearn-color-border-box);\n",
       "  border-radius: 0.25em;\n",
       "  box-sizing: border-box;\n",
       "  margin-bottom: 0.5em;\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-estimator.fitted {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "/* on hover */\n",
       "#sk-container-id-1 div.sk-estimator:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-estimator.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Specification for estimator info (e.g. \"i\" and \"?\") */\n",
       "\n",
       "/* Common style for \"i\" and \"?\" */\n",
       "\n",
       ".sk-estimator-doc-link,\n",
       "a:link.sk-estimator-doc-link,\n",
       "a:visited.sk-estimator-doc-link {\n",
       "  float: right;\n",
       "  font-size: smaller;\n",
       "  line-height: 1em;\n",
       "  font-family: monospace;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  border-radius: 1em;\n",
       "  height: 1em;\n",
       "  width: 1em;\n",
       "  text-decoration: none !important;\n",
       "  margin-left: 1ex;\n",
       "  /* unfitted */\n",
       "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-unfitted-level-1);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link.fitted,\n",
       "a:link.sk-estimator-doc-link.fitted,\n",
       "a:visited.sk-estimator-doc-link.fitted {\n",
       "  /* fitted */\n",
       "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-fitted-level-1);\n",
       "}\n",
       "\n",
       "/* On hover */\n",
       "div.sk-estimator:hover .sk-estimator-doc-link:hover,\n",
       ".sk-estimator-doc-link:hover,\n",
       "div.sk-label-container:hover .sk-estimator-doc-link:hover,\n",
       ".sk-estimator-doc-link:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "div.sk-estimator.fitted:hover .sk-estimator-doc-link.fitted:hover,\n",
       ".sk-estimator-doc-link.fitted:hover,\n",
       "div.sk-label-container:hover .sk-estimator-doc-link.fitted:hover,\n",
       ".sk-estimator-doc-link.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "/* Span, style for the box shown on hovering the info icon */\n",
       ".sk-estimator-doc-link span {\n",
       "  display: none;\n",
       "  z-index: 9999;\n",
       "  position: relative;\n",
       "  font-weight: normal;\n",
       "  right: .2ex;\n",
       "  padding: .5ex;\n",
       "  margin: .5ex;\n",
       "  width: min-content;\n",
       "  min-width: 20ex;\n",
       "  max-width: 50ex;\n",
       "  color: var(--sklearn-color-text);\n",
       "  box-shadow: 2pt 2pt 4pt #999;\n",
       "  /* unfitted */\n",
       "  background: var(--sklearn-color-unfitted-level-0);\n",
       "  border: .5pt solid var(--sklearn-color-unfitted-level-3);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link.fitted span {\n",
       "  /* fitted */\n",
       "  background: var(--sklearn-color-fitted-level-0);\n",
       "  border: var(--sklearn-color-fitted-level-3);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link:hover span {\n",
       "  display: block;\n",
       "}\n",
       "\n",
       "/* \"?\"-specific style due to the `<a>` HTML tag */\n",
       "\n",
       "#sk-container-id-1 a.estimator_doc_link {\n",
       "  float: right;\n",
       "  font-size: 1rem;\n",
       "  line-height: 1em;\n",
       "  font-family: monospace;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  border-radius: 1rem;\n",
       "  height: 1rem;\n",
       "  width: 1rem;\n",
       "  text-decoration: none;\n",
       "  /* unfitted */\n",
       "  color: var(--sklearn-color-unfitted-level-1);\n",
       "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 a.estimator_doc_link.fitted {\n",
       "  /* fitted */\n",
       "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-fitted-level-1);\n",
       "}\n",
       "\n",
       "/* On hover */\n",
       "#sk-container-id-1 a.estimator_doc_link:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 a.estimator_doc_link.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-3);\n",
       "}\n",
       "</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>VotingClassifier(estimators=[(&#x27;Cleveland Clinic&#x27;,\n",
       "                              RandomForestClassifier(random_state=12345)),\n",
       "                             (&#x27;Hungarian Inst. of Cardiology&#x27;,\n",
       "                              RandomForestClassifier(random_state=12345)),\n",
       "                             (&#x27;Univ. Hospitals Zurich and Basel&#x27;,\n",
       "                              RandomForestClassifier(random_state=12345)),\n",
       "                             (&#x27;V.A. Medical Center&#x27;,\n",
       "                              RandomForestClassifier(random_state=12345))],\n",
       "                 weights=[2, 1, 0.2, 0.5])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" ><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow fitted\">&nbsp;&nbsp;VotingClassifier<a class=\"sk-estimator-doc-link fitted\" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.5/modules/generated/sklearn.ensemble.VotingClassifier.html\">?<span>Documentation for VotingClassifier</span></a><span class=\"sk-estimator-doc-link fitted\">i<span>Fitted</span></span></label><div class=\"sk-toggleable__content fitted\"><pre>VotingClassifier(estimators=[(&#x27;Cleveland Clinic&#x27;,\n",
       "                              RandomForestClassifier(random_state=12345)),\n",
       "                             (&#x27;Hungarian Inst. of Cardiology&#x27;,\n",
       "                              RandomForestClassifier(random_state=12345)),\n",
       "                             (&#x27;Univ. Hospitals Zurich and Basel&#x27;,\n",
       "                              RandomForestClassifier(random_state=12345)),\n",
       "                             (&#x27;V.A. Medical Center&#x27;,\n",
       "                              RandomForestClassifier(random_state=12345))],\n",
       "                 weights=[2, 1, 0.2, 0.5])</pre></div> </div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label fitted sk-toggleable\"><label>Cleveland Clinic</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow fitted\">&nbsp;RandomForestClassifier<a class=\"sk-estimator-doc-link fitted\" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.5/modules/generated/sklearn.ensemble.RandomForestClassifier.html\">?<span>Documentation for RandomForestClassifier</span></a></label><div class=\"sk-toggleable__content fitted\"><pre>RandomForestClassifier(random_state=12345)</pre></div> </div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label fitted sk-toggleable\"><label>Hungarian Inst. of Cardiology</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow fitted\">&nbsp;RandomForestClassifier<a class=\"sk-estimator-doc-link fitted\" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.5/modules/generated/sklearn.ensemble.RandomForestClassifier.html\">?<span>Documentation for RandomForestClassifier</span></a></label><div class=\"sk-toggleable__content fitted\"><pre>RandomForestClassifier(random_state=12345)</pre></div> </div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label fitted sk-toggleable\"><label>Univ. Hospitals Zurich and Basel</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-4\" type=\"checkbox\" ><label for=\"sk-estimator-id-4\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow fitted\">&nbsp;RandomForestClassifier<a class=\"sk-estimator-doc-link fitted\" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.5/modules/generated/sklearn.ensemble.RandomForestClassifier.html\">?<span>Documentation for RandomForestClassifier</span></a></label><div class=\"sk-toggleable__content fitted\"><pre>RandomForestClassifier(random_state=12345)</pre></div> </div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label fitted sk-toggleable\"><label>V.A. Medical Center</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-5\" type=\"checkbox\" ><label for=\"sk-estimator-id-5\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow fitted\">&nbsp;RandomForestClassifier<a class=\"sk-estimator-doc-link fitted\" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.5/modules/generated/sklearn.ensemble.RandomForestClassifier.html\">?<span>Documentation for RandomForestClassifier</span></a></label><div class=\"sk-toggleable__content fitted\"><pre>RandomForestClassifier(random_state=12345)</pre></div> </div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "VotingClassifier(estimators=[('Cleveland Clinic',\n",
       "                              RandomForestClassifier(random_state=12345)),\n",
       "                             ('Hungarian Inst. of Cardiology',\n",
       "                              RandomForestClassifier(random_state=12345)),\n",
       "                             ('Univ. Hospitals Zurich and Basel',\n",
       "                              RandomForestClassifier(random_state=12345)),\n",
       "                             ('V.A. Medical Center',\n",
       "                              RandomForestClassifier(random_state=12345))],\n",
       "                 weights=[2, 1, 0.2, 0.5])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Explore \"Voting Classifier\" as Ensemble model strategy  \n",
    "from sklearn.ensemble import VotingClassifier\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "\n",
    "y = mock_data[\"num\"].map(lambda v: 0 if v == 0 else 1)\n",
    "\n",
    "# assign weights based on performance calculated in Experiment 03\n",
    "voting_model = VotingClassifier(estimators=[(name, model) for name, model in models.items()], \n",
    "                                weights=[2,1,0.2,0.5])\n",
    "\n",
    "# workaround to by-pass re-fit: https://stackoverflow.com/a/54610569\n",
    "voting_model.estimators_ = list(models.values())\n",
    "voting_model.le_ = LabelEncoder().fit(y)\n",
    "voting_model.classes_ = voting_model.le_.classes_\n",
    "\n",
    "voting_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b792ad39",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-0.06419407387663695\n",
      "[[ 2  9]\n",
      " [11 33]]\n"
     ]
    }
   ],
   "source": [
    "# DS/ML libraries\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import matthews_corrcoef as mcc_score\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "\n",
    "# ML Data preparation - same strategy as in model training\n",
    "def by_demographics(data):\n",
    "    sex = data[\"sex\"].map(lambda v: '0' if v == 0 else '1')\n",
    "    target = data[\"num\"].map(lambda v: '0' if v == 0 else '1')\n",
    "    return (sex+target).values\n",
    "\n",
    "\n",
    "# 1. get features and labels\n",
    "X = mock_data.drop(columns=[\"age\", \"sex\", \"num\"], axis=1)\n",
    "y = mock_data[\"num\"].map(lambda v: 0 if v == 0 else 1)\n",
    "# 2. partition data\n",
    "_, X_test, _, y_test = train_test_split(\n",
    "    X, y, random_state=12345, stratify=by_demographics(mock_data)\n",
    ")\n",
    "y_pred = voting_model.predict(X_test)\n",
    "print(mcc_score(y_test, y_pred))\n",
    "print(confusion_matrix(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2dcf3ab9-5efe-4b80-80b8-77073dd67bbd",
   "metadata": {},
   "source": [
    "## Step 3. Submit Experiment to each datasite"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9344b470-4ad8-4e25-894f-29c364ab7899",
   "metadata": {},
   "source": [
    "_Another evaluation experiment, this time using an `EnsembleVotingClassifier` model on each datasite!_\n",
    "\n",
    "> 🙋 **Note**: The code of the `evaluate` function we are going to create is **identical** to the one we used in our previous evaluation experiment (see [`03-Model-Evaluation-Experiment.ipynb`](./03-Model-Evaluation-Experiment.ipynb)).\n",
    ">\n",
    "> It has been replicated here for clarity - but a **better way** to handle this would've been to import the code from a module, and just change the input parameter of the `syft_function_single_use` decorator!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "cdb3fd07-67e6-4798-845f-4a84416e27b1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Datasite: Cleveland Clinic\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div class=\"alert-success\" style=\"padding:5px;\"><strong>SyftSuccess</strong>: <pre class=\"alert-success\" style=\"display:inline; font-family:inherit;\">Syft function 'evaluate' successfully created. To add a code request, please create a project using `project = syft.Project(...)`, then use command `project.create_code_request`.</pre></div><br/>"
      ],
      "text/plain": [
       "SyftSuccess: Syft function 'evaluate' successfully created. To add a code request, please create a project using `project = syft.Project(...)`, then use command `project.create_code_request`."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Datasite: Hungarian Inst. of Cardiology\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div class=\"alert-success\" style=\"padding:5px;\"><strong>SyftSuccess</strong>: <pre class=\"alert-success\" style=\"display:inline; font-family:inherit;\">Syft function 'evaluate' successfully created. To add a code request, please create a project using `project = syft.Project(...)`, then use command `project.create_code_request`.</pre></div><br/>"
      ],
      "text/plain": [
       "SyftSuccess: Syft function 'evaluate' successfully created. To add a code request, please create a project using `project = syft.Project(...)`, then use command `project.create_code_request`."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Datasite: Univ. Hospitals Zurich and Basel\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div class=\"alert-success\" style=\"padding:5px;\"><strong>SyftSuccess</strong>: <pre class=\"alert-success\" style=\"display:inline; font-family:inherit;\">Syft function 'evaluate' successfully created. To add a code request, please create a project using `project = syft.Project(...)`, then use command `project.create_code_request`.</pre></div><br/>"
      ],
      "text/plain": [
       "SyftSuccess: Syft function 'evaluate' successfully created. To add a code request, please create a project using `project = syft.Project(...)`, then use command `project.create_code_request`."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Datasite: V.A. Medical Center\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div class=\"alert-success\" style=\"padding:5px;\"><strong>SyftSuccess</strong>: <pre class=\"alert-success\" style=\"display:inline; font-family:inherit;\">Syft function 'evaluate' successfully created. To add a code request, please create a project using `project = syft.Project(...)`, then use command `project.create_code_request`.</pre></div><br/>"
      ],
      "text/plain": [
       "SyftSuccess: Syft function 'evaluate' successfully created. To add a code request, please create a project using `project = syft.Project(...)`, then use command `project.create_code_request`."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from utils import serialize_and_upload\n",
    "\n",
    "remote_models = {}\n",
    "for name, datasite in datasites.items():\n",
    "    print(f\"Datasite: {name}\")\n",
    "    # 1. Get data asset from datasite\n",
    "    data_asset = datasite.datasets[\"Heart Disease Dataset\"].assets[\"Heart Study Data\"]\n",
    "    # 1.1 Upload and serialise the Voting classifier to each Datasite\n",
    "    remote_voting_model = serialize_and_upload(model=voting_model, to=datasite)\n",
    "    remote_models[name] = remote_voting_model\n",
    "    \n",
    "    @sy.syft_function_single_use(data=data_asset, model=remote_voting_model)\n",
    "    def evaluate(data, model):\n",
    "        # DS/ML libraries\n",
    "        from sklearn.model_selection import train_test_split\n",
    "        from sklearn.metrics import matthews_corrcoef as mcc_score\n",
    "        from sklearn.metrics import confusion_matrix\n",
    "        import joblib  # to load serialised input model\n",
    "        \n",
    "        # ML Data preparation - same strategy as in model training\n",
    "        def by_demographics(data):\n",
    "            sex = data[\"sex\"].map(lambda v: '0' if v == 0 else '1')\n",
    "            target = data[\"num\"].map(lambda v: '0' if v == 0 else '1')\n",
    "            return (sex+target).values\n",
    "        \n",
    "        # 1. get features and labels\n",
    "        X = data.drop(columns=[\"age\", \"sex\", \"num\"], axis=1)\n",
    "        y = data[\"num\"].map(lambda v: 0 if v == 0 else 1)\n",
    "        # 2. partition data\n",
    "        _, X_test, _, y_test = train_test_split(\n",
    "            X, y, random_state=12345, stratify=by_demographics(data)\n",
    "        )\n",
    "        # 3. Get trained model\n",
    "        classifier = joblib.load(model)  # load model\n",
    "        # 4. Evaluate Metrics (MCC and Confusion Matrix)\n",
    "        y_pred = classifier.predict(X_test)\n",
    "        return mcc_score(y_test, y_pred), confusion_matrix(y_test, y_pred)\n",
    "    \n",
    "    ensemble_ml_project = sy.Project(\n",
    "        name=\"Evaluate Ensable Voting Classifier on Heart Study Data\",\n",
    "        description=\"\"\"This time I would like to evaluate the performance of an\n",
    "        Ensemble Voting Models, using pre-trained models on each datasite.\"\"\",\n",
    "        members=[datasite],\n",
    "    )\n",
    "    ensemble_ml_project.create_code_request(evaluate, datasite)\n",
    "    project = ensemble_ml_project.send()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "06e608be-ecf0-42b0-b41c-066b2adffc8e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Datasite: Cleveland Clinic\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div class=\"alert-success\" style=\"padding:5px;\"><strong>SyftSuccess</strong>: <pre class=\"alert-success\" style=\"display:inline; font-family:inherit;\"> approved</pre></div><br/>"
      ],
      "text/plain": [
       "SyftSuccess: <class 'syft.service.code.user_code.UserCodeStatusCollection'> approved"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Datasite: Hungarian Inst. of Cardiology\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div class=\"alert-success\" style=\"padding:5px;\"><strong>SyftSuccess</strong>: <pre class=\"alert-success\" style=\"display:inline; font-family:inherit;\"> approved</pre></div><br/>"
      ],
      "text/plain": [
       "SyftSuccess: <class 'syft.service.code.user_code.UserCodeStatusCollection'> approved"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Datasite: Univ. Hospitals Zurich and Basel\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div class=\"alert-success\" style=\"padding:5px;\"><strong>SyftSuccess</strong>: <pre class=\"alert-success\" style=\"display:inline; font-family:inherit;\"> approved</pre></div><br/>"
      ],
      "text/plain": [
       "SyftSuccess: <class 'syft.service.code.user_code.UserCodeStatusCollection'> approved"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Datasite: V.A. Medical Center\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div class=\"alert-success\" style=\"padding:5px;\"><strong>SyftSuccess</strong>: <pre class=\"alert-success\" style=\"display:inline; font-family:inherit;\"> approved</pre></div><br/>"
      ],
      "text/plain": [
       "SyftSuccess: <class 'syft.service.code.user_code.UserCodeStatusCollection'> approved"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from utils import check_status_last_code_requests\n",
    "\n",
    "check_status_last_code_requests(datasites)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1d6ecd12-70e6-4a0d-b497-d0e2fbb7dcac",
   "metadata": {},
   "source": [
    "## Step 4. Evaluate Models on all datasites"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "4dcb4799-1f90-4b51-ba23-45934ab87c76",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Datasite: Cleveland Clinic\n",
      "Datasite: Hungarian Inst. of Cardiology\n",
      "Datasite: Univ. Hospitals Zurich and Basel\n",
      "Datasite: V.A. Medical Center\n"
     ]
    }
   ],
   "source": [
    "mcc_scores, confusion_matrices = {}, {}\n",
    "for name, datasite in datasites.items():\n",
    "    print(f\"Datasite: {name}\")\n",
    "    data_asset = datasite.datasets[\"Heart Disease Dataset\"].assets[\"Heart Study Data\"]\n",
    "    remote_model = remote_models[name]\n",
    "    results = datasite.code.evaluate(data=data_asset, model=remote_model).get_from(datasite)\n",
    "    mcc_scores[name], confusion_matrices[name] = results"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b12f2cf-ab2a-4e74-a4ce-987fad66488d",
   "metadata": {},
   "source": [
    "**Note**: \n",
    "MCC Classification report of the `RandomForest` classifiers trained independently on each Datasites\n",
    "\n",
    "```\n",
    "'Cleveland Clinic': 0.766,\n",
    "'Hungarian Inst. of Cardiology': 0.46,\n",
    "'Univ. Hospitals Zurich and Basel': 0.0,\n",
    "'V.A. Medical Center': 0.258}\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d0600ae7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cleveland Clinic: 0.766\n",
      "Hungarian Inst. of Cardiology: 0.583\n",
      "Univ. Hospitals Zurich and Basel: 0.330\n",
      "V.A. Medical Center: 0.308\n"
     ]
    }
   ],
   "source": [
    "for name, score in mcc_scores.items():\n",
    "    print(f\"{name}: {score:.3f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "57142cd3-775f-40ef-a2b0-6212009cec50",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x800 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "from utils import plot_all_confusion_matrices\n",
    "\n",
    "plot_all_confusion_matrices(confusion_matrices)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0bfbbce0-923d-42c8-a495-d2576576ffdb",
   "metadata": {},
   "source": [
    "## Conclusions 🎉"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3a4cb45f-abdf-45a2-8771-7413c28a3e42",
   "metadata": {},
   "source": [
    "Using an Ensemble of all models, we have been able to _improve_ the results of our classifier on **all** the four datasets by _indirectly_ incorporating the knowledge gathered from each (non-public) dataset, independently!\n",
    "\n",
    "And we did this by **never, ever** _seeing_ the true private data stored in each datasite!"
   ]
  }
 ],
 "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": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}