tutorials/automl/tutorial_06_automl_multiple_imputation_example.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "623a0448",
   "metadata": {},
   "source": [
    "# Tutorial: Simulating multiple imputation(MICE) using AutoPrognosis\n",
    "\n",
    "Welcome to the classification AutoML tutorial!\n",
    "\n",
    "This tutorial will show how to use AutoPrognosis and multiple imputation to learn a model for datasets with missing data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8d3dc094",
   "metadata": {},
   "outputs": [],
   "source": [
    "# stdlib\n",
    "import json\n",
    "import sys\n",
    "import warnings\n",
    "\n",
    "# third party\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "# autoprognosis absolute\n",
    "import autoprognosis.logger as log\n",
    "from autoprognosis.studies.classifiers import ClassifierStudy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "15ec7b75",
   "metadata": {},
   "outputs": [],
   "source": [
    "log.add(sink=sys.stderr, level=\"INFO\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2f0827f5",
   "metadata": {},
   "source": [
    "## Load toy dataset\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c45ad99e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# stdlib\n",
    "from pathlib import Path\n",
    "\n",
    "\n",
    "def get_dataset() -> pd.DataFrame:\n",
    "    Path(\"data\").mkdir(parents=True, exist_ok=True)\n",
    "    bkp_file = Path(\"data\") / \"anneal.csv\"\n",
    "\n",
    "    if bkp_file.exists():\n",
    "        return pd.read_csv(bkp_file)\n",
    "\n",
    "    df = pd.read_csv(\n",
    "        \"https://archive.ics.uci.edu/ml/machine-learning-databases/annealing/anneal.data\",\n",
    "        header=None,\n",
    "    )\n",
    "    df.to_csv(bkp_file, index=None)\n",
    "\n",
    "    return df\n",
    "\n",
    "\n",
    "df = get_dataset()\n",
    "\n",
    "df = df.replace(\"?\", np.nan)\n",
    "\n",
    "X = df.drop(columns=[df.columns[-1]])\n",
    "y = df[df.columns[-1]]\n",
    "\n",
    "X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fe8bcb2c",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = X.copy()\n",
    "dataset[\"target\"] = y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "695eb4cf",
   "metadata": {},
   "outputs": [],
   "source": [
    "for col in X.columns:\n",
    "    if X[col].isna().sum() == 0:\n",
    "        continue\n",
    "\n",
    "    col_type = \"categorical\" if len(X[col].unique()) < 10 else \"cont\"\n",
    "    print(\n",
    "        f\"NaNs ratio in col = {col} col_type = {col_type} miss ratio = {X[col].isna().sum() / len(X[col])}\"\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "aea15e48",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "300dccdf",
   "metadata": {},
   "outputs": [],
   "source": [
    "# List available classifiers\n",
    "\n",
    "# autoprognosis absolute\n",
    "from autoprognosis.plugins.prediction import Classifiers\n",
    "\n",
    "Classifiers().list_available()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "343a2ff6",
   "metadata": {},
   "source": [
    "## Search model with the ICE imputer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "69b0146a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# stdlib\n",
    "from pathlib import Path\n",
    "\n",
    "workspace = Path(\"workspace\")\n",
    "workspace.mkdir(parents=True, exist_ok=True)\n",
    "\n",
    "study_name = \"test_classification_studies_mice\"\n",
    "\n",
    "study = ClassifierStudy(\n",
    "    study_name=study_name,\n",
    "    dataset=dataset,\n",
    "    target=\"target\",\n",
    "    imputers=[\n",
    "        \"ice\"\n",
    "    ],  # Using chained equations. Can use it for \"missforest\" or \"hyperimpute\" plugins as well.\n",
    "    num_iter=10,  # DELETE THIS LINE FOR BETTER RESULTS.\n",
    "    num_study_iter=1,  # DELETE THIS LINE FOR BETTER RESULTS.\n",
    "    classifiers=[\"logistic_regression\", \"lda\"],  # DELETE THIS LINE FOR BETTER RESULTS.\n",
    "    workspace=workspace,\n",
    ")\n",
    "study.run()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f76a0865",
   "metadata": {},
   "source": [
    "## Train the model template using multiple random seeds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2903d9b5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# autoprognosis absolute\n",
    "from autoprognosis.plugins.imputers import Imputers\n",
    "from autoprognosis.utils.serialization import load_model_from_file\n",
    "\n",
    "model_path = workspace / study_name / \"model.p\"\n",
    "\n",
    "model = load_model_from_file(model_path)\n",
    "\n",
    "model.name()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9b298666",
   "metadata": {},
   "outputs": [],
   "source": [
    "# autoprognosis absolute\n",
    "from autoprognosis.utils.distributions import enable_reproducible_results\n",
    "from autoprognosis.utils.tester import evaluate_estimator_multiple_seeds\n",
    "\n",
    "score = evaluate_estimator_multiple_seeds(model, X, y, seeds=list(range(5)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "539ef326",
   "metadata": {},
   "outputs": [],
   "source": [
    "score"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d8f69a68",
   "metadata": {},
   "source": [
    "## Congratulations!\n",
    "\n",
    "Congratulations on completing this notebook tutorial! If you enjoyed this and would like to join the movement towards Machine learning and AI for medicine, you can do so in the following ways!\n",
    "\n",
    "### Star AutoPrognosis on GitHub\n",
    "\n",
    "The easiest way to help our community is just by starring the Repos! This helps raise awareness of the tools we're building.\n",
    "\n",
    "- [Star AutoPrognosis](https://github.com/vanderschaarlab/autoprognosis)\n",
    "- [Star HyperImpute](https://github.com/vanderschaarlab/hyperimpute)\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": "3.9.15"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}