01b-cross_validation.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cross-validation for PU and SVM models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "use_extracted_data = False\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "env: PYTHONWARNINGS=ignore::FutureWarning\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import warnings\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from pulearn.bagging import BaggingPuClassifier\n",
    "from sklearn.base import clone\n",
    "from sklearn.metrics import f1_score\n",
    "from sklearn.model_selection import (\n",
    "    RepeatedStratifiedKFold,\n",
    "    train_test_split,\n",
    ")\n",
    "from sklearn.preprocessing import RobustScaler\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "from lib.check_files import check_prepared_data\n",
    "from lib.cv import perform_cv\n",
    "from lib.pu import (\n",
    "    BASE_MODELS,\n",
    "    PU_PARAMS,\n",
    "    get_xy,\n",
    "    downsample_unlabelled,\n",
    ")\n",
    "\n",
    "# Suppress FutureWarning for some versions of Scikit-learn\n",
    "%env PYTHONWARNINGS=ignore::FutureWarning\n",
    "\n",
    "warnings.simplefilter(\"ignore\", FutureWarning)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Random seed for reproducibility\n",
    "random_seed = 1234\n",
    "\n",
    "# Number of jobs used to train models\n",
    "n_jobs = 8\n",
    "\n",
    "rf_params = {\"random_state\": random_seed}\n",
    "pu_params = {\n",
    "    \"n_jobs\": n_jobs,\n",
    "    \"random_state\": random_seed,\n",
    "    **PU_PARAMS\n",
    "}\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load and process training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "label\n",
       "negative      392\n",
       "positive      190\n",
       "unlabelled    190\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "if use_extracted_data:\n",
    "    data_dir = \"extracted_data\"\n",
    "else:\n",
    "    data_dir = \"prepared_data\"\n",
    "    check_prepared_data(data_dir, verbose=True)\n",
    "\n",
    "figures_dir = \"supplementary_figures\"\n",
    "os.makedirs(figures_dir, exist_ok=True)\n",
    "\n",
    "data_filename = os.path.join(data_dir, \"training_data.csv\")\n",
    "data = pd.read_csv(data_filename)\n",
    "\n",
    "# Restrict unlabelled training data to the Americas\n",
    "data = data[data[\"region\"].isin({\"NAm\", \"SAm\"})]\n",
    "\n",
    "data = downsample_unlabelled(data, random_state=random_seed)\n",
    "\n",
    "data.groupby(\"label\").size()\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Perform cross-validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "cv = RepeatedStratifiedKFold(n_splits=5, n_repeats=5, random_state=random_seed)\n",
    "# Scaling not necessary for RF, but kept for consistency\n",
    "scaler = RobustScaler()\n",
    "\n",
    "thresh_pu = \"auto\"\n",
    "base_model = clone(BASE_MODELS[\"randomforest\"])\n",
    "base_model.set_params(**rf_params)\n",
    "model_pu = BaggingPuClassifier(base_model, **pu_params)\n",
    "model_pu = make_pipeline(clone(scaler), model_pu)\n",
    "results_pu = perform_cv(\n",
    "    clf=model_pu,\n",
    "    data=data,\n",
    "    cv=cv,\n",
    "    thresh=thresh_pu,\n",
    "    random_state=random_seed,\n",
    "    pu=True,\n",
    "    verbose=False,\n",
    ")\n",
    "results_pu = pd.DataFrame(results_pu)\n",
    "\n",
    "thresh_svm = \"auto\"\n",
    "model_svm = SVC(kernel=\"rbf\", probability=True, random_state=random_seed)\n",
    "model_svm = make_pipeline(clone(scaler), model_svm)\n",
    "\n",
    "results_svm = perform_cv(\n",
    "    clf=model_svm,\n",
    "    data=data,\n",
    "    cv=cv,\n",
    "    thresh=thresh_svm,\n",
    "    random_state=random_seed,\n",
    "    pu=False,\n",
    "    verbose=False,\n",
    ")\n",
    "results_svm = pd.DataFrame(results_svm)\n",
    "\n",
    "results_combined = pd.concat(\n",
    "    (\n",
    "        results_pu.assign(method=\"PU\"),\n",
    "        results_svm.assign(method=\"SVM\"),\n",
    "    ),\n",
    "    ignore_index=True,\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th>method</th>\n",
       "      <th colspan=\"3\" halign=\"left\">PU</th>\n",
       "      <th colspan=\"3\" halign=\"left\">SVM</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>region</th>\n",
       "      <th>All</th>\n",
       "      <th>NAm</th>\n",
       "      <th>SAm</th>\n",
       "      <th>All</th>\n",
       "      <th>NAm</th>\n",
       "      <th>SAm</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>roc_auc</th>\n",
       "      <td>0.954836</td>\n",
       "      <td>0.979391</td>\n",
       "      <td>0.881470</td>\n",
       "      <td>0.983288</td>\n",
       "      <td>0.948861</td>\n",
       "      <td>0.999055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>average_precision</th>\n",
       "      <td>0.960759</td>\n",
       "      <td>0.995109</td>\n",
       "      <td>0.829300</td>\n",
       "      <td>0.975712</td>\n",
       "      <td>0.981661</td>\n",
       "      <td>0.992687</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>balanced_accuracy</th>\n",
       "      <td>0.952199</td>\n",
       "      <td>0.955592</td>\n",
       "      <td>0.891653</td>\n",
       "      <td>0.968642</td>\n",
       "      <td>0.921211</td>\n",
       "      <td>0.988849</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>accuracy</th>\n",
       "      <td>0.966658</td>\n",
       "      <td>0.948105</td>\n",
       "      <td>0.976229</td>\n",
       "      <td>0.974574</td>\n",
       "      <td>0.938824</td>\n",
       "      <td>0.993795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1</th>\n",
       "      <td>0.946314</td>\n",
       "      <td>0.965160</td>\n",
       "      <td>0.851422</td>\n",
       "      <td>0.960413</td>\n",
       "      <td>0.959703</td>\n",
       "      <td>0.967959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n_test</th>\n",
       "      <td>116.400000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>76.400000</td>\n",
       "      <td>116.400000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>76.400000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n_train</th>\n",
       "      <td>273.600000</td>\n",
       "      <td>273.600000</td>\n",
       "      <td>273.600000</td>\n",
       "      <td>465.600000</td>\n",
       "      <td>465.600000</td>\n",
       "      <td>465.600000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time_fit</th>\n",
       "      <td>2.021926</td>\n",
       "      <td>2.021926</td>\n",
       "      <td>2.021926</td>\n",
       "      <td>0.022801</td>\n",
       "      <td>0.022801</td>\n",
       "      <td>0.022801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time_predict</th>\n",
       "      <td>1.717679</td>\n",
       "      <td>0.848375</td>\n",
       "      <td>0.869304</td>\n",
       "      <td>0.002014</td>\n",
       "      <td>0.000852</td>\n",
       "      <td>0.001161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>prob_thresh</th>\n",
       "      <td>0.554343</td>\n",
       "      <td>0.554343</td>\n",
       "      <td>0.554343</td>\n",
       "      <td>0.367677</td>\n",
       "      <td>0.367677</td>\n",
       "      <td>0.367677</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "method                     PU                                 SVM              \\\n",
       "region                    All         NAm         SAm         All         NAm   \n",
       "roc_auc              0.954836    0.979391    0.881470    0.983288    0.948861   \n",
       "average_precision    0.960759    0.995109    0.829300    0.975712    0.981661   \n",
       "balanced_accuracy    0.952199    0.955592    0.891653    0.968642    0.921211   \n",
       "accuracy             0.966658    0.948105    0.976229    0.974574    0.938824   \n",
       "f1                   0.946314    0.965160    0.851422    0.960413    0.959703   \n",
       "n_test             116.400000   40.000000   76.400000  116.400000   40.000000   \n",
       "n_train            273.600000  273.600000  273.600000  465.600000  465.600000   \n",
       "time_fit             2.021926    2.021926    2.021926    0.022801    0.022801   \n",
       "time_predict         1.717679    0.848375    0.869304    0.002014    0.000852   \n",
       "prob_thresh          0.554343    0.554343    0.554343    0.367677    0.367677   \n",
       "\n",
       "method                         \n",
       "region                    SAm  \n",
       "roc_auc              0.999055  \n",
       "average_precision    0.992687  \n",
       "balanced_accuracy    0.988849  \n",
       "accuracy             0.993795  \n",
       "f1                   0.967959  \n",
       "n_test              76.400000  \n",
       "n_train            465.600000  \n",
       "time_fit             0.022801  \n",
       "time_predict         0.001161  \n",
       "prob_thresh          0.367677  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_combined.groupby([\"method\", \"region\"]).mean(numeric_only=True).T\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Performance varies by region"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "set    region  label     \n",
       "test   NAm     negative        9\n",
       "               positive       31\n",
       "       SAm     negative       70\n",
       "               positive        7\n",
       "train  NAm     negative       36\n",
       "               positive      124\n",
       "               unlabelled     87\n",
       "       SAm     negative      277\n",
       "               positive       28\n",
       "               unlabelled     65\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stratify = (\n",
    "    data[\"region\"]\n",
    "    + \"_\"\n",
    "    + data[\"label\"]\n",
    ")\n",
    "data_train, data_test = train_test_split(\n",
    "    data,\n",
    "    test_size=0.2,\n",
    "    random_state=random_seed,\n",
    "    shuffle=True,\n",
    "    stratify=stratify,\n",
    ")\n",
    "data_test = data_test[data_test[\"label\"].isin({\"positive\", \"negative\"})]\n",
    "indices_NA = np.where(data_test[\"region\"] == \"NAm\")[0]\n",
    "indices_SA = np.where(data_test[\"region\"] == \"SAm\")[0]\n",
    "\n",
    "train_pu = data_train[\n",
    "    data_train[\"label\"].isin({\"positive\", \"unlabeled\", \"unlabelled\"})\n",
    "]\n",
    "x_train_pu, y_train_pu = get_xy(train_pu)\n",
    "train_svm = data_train[\n",
    "    data_train[\"label\"].isin({\"positive\", \"negative\"})\n",
    "]\n",
    "x_train_svm, y_train_svm = get_xy(train_svm)\n",
    "\n",
    "x_test, y_test = get_xy(data_test)\n",
    "\n",
    "model_pu = clone(model_pu)\n",
    "model_pu.fit(x_train_pu, y_train_pu)\n",
    "\n",
    "model_svm = clone(model_svm)\n",
    "model_svm.fit(x_train_svm, y_train_svm)\n",
    "\n",
    "probs_pu = model_pu.predict_proba(x_test)[:, 1]\n",
    "probs_pu_NA = probs_pu[indices_NA]\n",
    "probs_pu_SA = probs_pu[indices_SA]\n",
    "\n",
    "probs_svm = model_svm.predict_proba(x_test)[:, 1]\n",
    "probs_svm_NA = probs_svm[indices_NA]\n",
    "probs_svm_SA = probs_svm[indices_SA]\n",
    "\n",
    "probs_dict = {\n",
    "    \"All\": {\"PU\": probs_pu, \"SVM\": probs_svm},\n",
    "    \"NA\": {\"PU\": probs_pu_NA, \"SVM\": probs_svm_NA},\n",
    "    \"SA\": {\"PU\": probs_pu_SA, \"SVM\": probs_svm_SA},\n",
    "}\n",
    "kwargs_dict = {\n",
    "    \"PU\": {\"color\": \"blue\", \"linestyle\": \"solid\"},\n",
    "    \"SVM\": {\"color\": \"orange\", \"linestyle\": \"dashed\"},\n",
    "}\n",
    "\n",
    "pd.concat(\n",
    "    (\n",
    "        data_train.assign(set=\"train\"),\n",
    "        data_test.assign(set=\"test\"),\n",
    "    )\n",
    ").groupby([\"set\", \"region\", \"label\"]).size()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x1600 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figure_filename = os.path.join(figures_dir, \"cv_boxplots\")\n",
    "\n",
    "regions = (\"All\", \"NAm\", \"SAm\")\n",
    "region_format = {\n",
    "    \"NAm\": \"North America\",\n",
    "    \"SAm\": \"South America\",\n",
    "}\n",
    "metric_format = {\n",
    "    \"roc_auc\": \"ROC AUC\",\n",
    "    \"average_precision\": \"Average\\nprecision\",\n",
    "    \"balanced_accuracy\": \"Balanced\\naccuracy\",\n",
    "    \"accuracy\": \"Accuracy\",\n",
    "    \"f1\": \"F1\",\n",
    "}\n",
    "drop = [\n",
    "    \"n_train\",\n",
    "    \"time_fit\",\n",
    "    \"n_test\",\n",
    "    \"time_predict\",\n",
    "    \"prob_thresh\",\n",
    "    \"method\",\n",
    "    \"region\",\n",
    "]\n",
    "\n",
    "fig, axs = plt.subplots(3, 2, figsize=(12, 16), sharey=\"row\")\n",
    "for i, region in enumerate(regions):\n",
    "    results_region = results_combined[results_combined[\"region\"] == region]\n",
    "    for j, (method, results_method) in enumerate(\n",
    "        results_region.groupby(\"method\")\n",
    "    ):\n",
    "        ax = axs[i, j]\n",
    "        results_method = results_method.drop(columns=drop, errors=\"ignore\")\n",
    "        results_method.boxplot(ax=axs[i, j], rot=0)\n",
    "        title = f\"{method} ({region_format.get(region, region)})\"\n",
    "        ax.set_title(title)\n",
    "        ax.set_xticklabels(\n",
    "            [\n",
    "                metric_format.get(i.get_text(), i.get_text())\n",
    "                for i in ax.get_xticklabels()\n",
    "            ]\n",
    "        )\n",
    "\n",
    "for ext in (\".pdf\", \".png\"):\n",
    "    fig.savefig(figure_filename + ext, dpi=350, bbox_inches=\"tight\")\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Determining the optimal probability threshold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PU: optimal threshold = 0.54\n",
      "SVM: optimal threshold = 0.58\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figure_filename = os.path.join(figures_dir, \"probability_threshold\")\n",
    "\n",
    "score_func = f1_score\n",
    "thresholds = np.linspace(0, 1, 100)\n",
    "\n",
    "fig, ax = plt.subplots(1, 1)\n",
    "for model in (\"PU\", \"SVM\"):\n",
    "    probs = probs_dict[\"All\"][model]\n",
    "    scores = np.zeros_like(thresholds)\n",
    "    for i, threshold in enumerate(thresholds):\n",
    "        score = score_func(y_test, probs >= threshold)\n",
    "        scores[i] = score\n",
    "    ax.plot(thresholds, scores, label=model, **(kwargs_dict[model]))\n",
    "    opt_value = np.median(thresholds[scores == np.nanmax(scores)])\n",
    "    ax.axvline(opt_value, 0, 1, **(kwargs_dict[model]), alpha=0.5)\n",
    "    print(f\"{model}: optimal threshold = {opt_value:0.2f}\")\n",
    "\n",
    "ax.set_xlim(-0.01, 1.01)\n",
    "ax.grid(\n",
    "    linestyle=\"dotted\",\n",
    "    color=\"grey\",\n",
    "    alpha=0.5,\n",
    ")\n",
    "ax.legend()\n",
    "ax.set_xlabel(\"Probability threshold\")\n",
    "ylabel = r\"$\\mathrm{F_1}$ score\"\n",
    "ax.set_ylabel(ylabel)\n",
    "ax.set_title(f\"{ylabel} v. probability threshold\")\n",
    "\n",
    "for ext in (\".pdf\", \".png\"):\n",
    "    fig.savefig(figure_filename + ext, dpi=350, bbox_inches=\"tight\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "datamining",
   "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": "3.11.5"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}