classfication_imagenette_imagewoof.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Global Configuration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from tqdm import tqdm\n",
    "import pickle\n",
    "import os\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "import tensorflow as tf\n",
    "\n",
    "from deel.puncc.api.prediction import BasePredictor\n",
    "from deel.puncc.classification import APS, RAPS\n",
    "\n",
    "# Choose CPU or GPU\n",
    "os.environ[\"CUDA_VISIBLE_DEVICES\"] = \"\"  # \"0\"\n",
    "\n",
    "# Load data\n",
    "with open(\"data/imagenet_dict.txt\") as f:\n",
    "    data = f.read()\n",
    "    imagenet_catalogue = eval(data)\n",
    "\n",
    "# Load ResNet-50 with weights\n",
    "model = tf.keras.applications.ResNet50(weights=\"imagenet\")\n",
    "model.compile(\n",
    "    loss=\"categorical_crossentropy\", optimizer=\"rmsprop\", metrics=[\"accuracy\"]\n",
    ")\n",
    "\n",
    "# Wrap ResNet-50 in a BasePredictor and indicate that the model is already trained\n",
    "resnet50_predictor = BasePredictor(model, is_trained=True)\n",
    "\n",
    "\n",
    "# Source: https://docs.fast.ai/tutorial.imagenette.html\n",
    "directory_imagenette_labels_dict = {\n",
    "    \"n01440764\": \"tench\",\n",
    "    \"n02102040\": \"English springer\",\n",
    "    \"n02979186\": \"cassette player\",\n",
    "    \"n03000684\": \"chain saw\",\n",
    "    \"n03028079\": \"church\",\n",
    "    \"n03394916\": \"French horn\",\n",
    "    \"n03417042\": \"garbage truck\",\n",
    "    \"n03425413\": \"gas pump\",\n",
    "    \"n03445777\": \"golf ball\",\n",
    "    \"n03888257\": \"parachute\",\n",
    "}\n",
    "\n",
    "# Source: https://walkwithfastai.com/DBlock_Summary\n",
    "directory_dogbreed_dict = {\n",
    "    \"n02086240\": \"Shih-Tzu\",\n",
    "    \"n02087394\": \"Rhodesian ridgeback\",\n",
    "    \"n02088364\": \"Beagle\",\n",
    "    \"n02089973\": \"English foxhound\",\n",
    "    \"n02093754\": \"Australian terrier\",\n",
    "    \"n02096294\": \"Border terrier\",\n",
    "    \"n02099601\": \"Golden retriever\",\n",
    "    \"n02105641\": \"Old English sheepdog\",\n",
    "    \"n02111889\": \"Samoyed\",\n",
    "    \"n02115641\": \"Dingo\",\n",
    "}\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data loading and preparation: Imagenette "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Imagenette is downloadable from https://s3.amazonaws.com/fast-ai-imageclas/imagenette2-320.tgz\n",
    "# Extract in data folder\n",
    "\n",
    "\n",
    "# create a new generator\n",
    "imagegen = ImageDataGenerator()\n",
    "\n",
    "# Number of images to load per iteration\n",
    "batch_size = 2048\n",
    "\n",
    "# Takes the path to imagenette to generates train and validation batches\n",
    "imagenette_train = imagegen.flow_from_directory(\n",
    "    \"data/imagenette/train\",\n",
    "    class_mode=\"binary\",\n",
    "    shuffle=True,\n",
    "    seed=0,\n",
    "    batch_size=batch_size,\n",
    "    target_size=(224, 224),\n",
    ")\n",
    "imagenette_val = imagegen.flow_from_directory(\n",
    "    \"data/imagenette/val\",\n",
    "    class_mode=\"binary\",\n",
    "    shuffle=True,\n",
    "    seed=0,\n",
    "    batch_size=batch_size,\n",
    "    target_size=(224, 224),\n",
    ")\n",
    "\n",
    "# Dictionary imagenette label => file name\n",
    "imagenette_label_file_catalogue = {\n",
    "    v: k for k, v in imagenette_val.class_indices.items()\n",
    "}\n",
    "\n",
    "# Dictionary imagenette label => name\n",
    "imagenette_catalogue = {\n",
    "    k: directory_imagenette_labels_dict[v]\n",
    "    for k, v in imagenette_label_file_catalogue.items()\n",
    "}\n",
    "\n",
    "# Dictionary imagenette label => imagenet label\n",
    "imagenette_imagenet_label_map = {}\n",
    "for imagenette_key, imagenette_value in imagenette_catalogue.items():\n",
    "    imagenet_key = [\n",
    "        k for k, v in imagenet_catalogue.items() if imagenette_value in v\n",
    "    ]\n",
    "    imagenette_imagenet_label_map[imagenette_key] = imagenet_key[0]\n",
    "\n",
    "\n",
    "# Function to translate labels encoded from Imagenette to Imagenet\n",
    "def nette_to_net_labels(labels):\n",
    "    return np.array([imagenette_imagenet_label_map[label] for label in labels])\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conformal prediction on Imagenette with ResNet-50"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The predictor is already trained, we will skip model fitting\n",
    "# by setting argument `train` to False\n",
    "imagenette_aps = APS(resnet50_predictor, train=False)\n",
    "imagenette_raps = RAPS(resnet50_predictor, train=False, lambd=0.01, k_reg=6)\n",
    "\n",
    "# fit on Imagenette\n",
    "for images, labels in imagenette_train:\n",
    "    X_calib = images\n",
    "    y_calib = nette_to_net_labels(labels)\n",
    "    # Provide only calibration set as model is already trained\n",
    "    imagenette_aps.fit(X_calib=X_calib, y_calib=y_calib)\n",
    "    imagenette_raps.fit(X_calib=X_calib, y_calib=y_calib)\n",
    "    break\n",
    "\n",
    "# Values of alpha to experiment with\n",
    "alphas = np.arange(0.1, 1, 0.1)\n",
    "\n",
    "# Test on imagenette\n",
    "results = []\n",
    "for images, labels in imagenette_val:\n",
    "    X_test = images\n",
    "    y_test = nette_to_net_labels(labels)\n",
    "\n",
    "    # Build prediction sets for APS and RAPS with different values of alpha\n",
    "    # and compute metrics\n",
    "    for alpha in tqdm(alphas):\n",
    "        # Prediction\n",
    "        y_pred_aps, set_pred_aps = imagenette_aps.predict(X_test, alpha=alpha)\n",
    "        y_pred_raps, set_pred_raps = imagenette_raps.predict(\n",
    "            X_test, alpha=alpha\n",
    "        )\n",
    "\n",
    "        # Compute average coverage and size\n",
    "        coverage_aps = classification_mean_coverage(y_test, set_pred_aps)\n",
    "        coverage_raps = classification_mean_coverage(y_test, set_pred_raps)\n",
    "        mean_size_aps = classification_mean_size(set_pred_aps)\n",
    "        mean_size_raps = classification_mean_size(set_pred_raps)\n",
    "\n",
    "        # Append iteration results\n",
    "        results.append([\"APS Imagenette\", alpha, coverage_aps, mean_size_aps])\n",
    "        results.append(\n",
    "            [\"RAPS Imagenette\", alpha, coverage_raps, mean_size_raps]\n",
    "        )\n",
    "    break\n",
    "\n",
    "with open(\"results/results_classification_imagenette.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(results, output_file)\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data loading and preparation: Imagewoof "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Imagewoof is downloadable from https://s3.amazonaws.com/fast-ai-imageclas/imagewoof2-320.tgz\n",
    "# Extract in data folder\n",
    "\n",
    "# create a new generator\n",
    "imagegen = ImageDataGenerator()\n",
    "\n",
    "# Number of images to load per iteration\n",
    "batch_size = 2048\n",
    "\n",
    "# Takes the path to imagewoof directory to generates train and validation batches\n",
    "imagewoof_train = imagegen.flow_from_directory(\n",
    "    \"data/imagewoof/train\",\n",
    "    class_mode=\"binary\",\n",
    "    shuffle=True,\n",
    "    seed=0,\n",
    "    batch_size=batch_size,\n",
    "    target_size=(224, 224),\n",
    ")\n",
    "imagewoof_val = imagegen.flow_from_directory(\n",
    "    \"data/imagewoof/val\",\n",
    "    class_mode=\"binary\",\n",
    "    shuffle=True,\n",
    "    seed=0,\n",
    "    batch_size=batch_size,\n",
    "    target_size=(224, 224),\n",
    ")\n",
    "\n",
    "# Dictionary imagewoof label => file name\n",
    "imagewoof_label_file_catalogue = {\n",
    "    v: k for k, v in imagewoof_val.class_indices.items()\n",
    "}\n",
    "\n",
    "# Dictionary imagewoof label => dogbreed\n",
    "imagewoof_catalogue = {\n",
    "    k: directory_dogbreed_dict[v]\n",
    "    for k, v in imagewoof_label_file_catalogue.items()\n",
    "}\n",
    "\n",
    "# Dictionary imagewoof label => imagenet label\n",
    "imagewoof_imagenet_label_map = {}\n",
    "for imagewoof_key, imagewoof_value in imagewoof_catalogue.items():\n",
    "    imagenet_key = [\n",
    "        k for k, v in imagenet_catalogue.items() if imagewoof_value in v\n",
    "    ]\n",
    "    imagewoof_imagenet_label_map[imagewoof_key] = imagenet_key[0]\n",
    "\n",
    "\n",
    "# Function to translate labels encoded from Imagewoof to Imagenet\n",
    "def woof_to_net_labels(labels):\n",
    "    return np.array([imagewoof_imagenet_label_map[label] for label in labels])\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conformal prediction on Imagewoof with ResNet-50"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The predictor is already trained, we will skip model fitting\n",
    "# by setting argument `train` to False\n",
    "imagewoof_aps = APS(resnet50_predictor, train=False)\n",
    "imagewoof_raps = RAPS(resnet50_predictor, train=False, lambd=0.01, k_reg=6)\n",
    "\n",
    "# Fit APS and RAPS on Imagewoof\n",
    "for images, labels in imagewoof_train:\n",
    "    X_calib = images\n",
    "    y_calib = woof_to_net_labels(labels)\n",
    "    # Provide only calibration set as model is already trained\n",
    "    imagewoof_aps.fit(X_calib=X_calib, y_calib=y_calib)\n",
    "    imagewoof_raps.fit(X_calib=X_calib, y_calib=y_calib)\n",
    "    break\n",
    "\n",
    "# Values of alpha to experiment with\n",
    "alphas = np.arange(0.1, 1, 0.1)\n",
    "\n",
    "# Test on imagewoof\n",
    "results = []\n",
    "for images, labels in tqdm(imagewoof_val):\n",
    "    X_test = images\n",
    "    y_test = woof_to_net_labels(labels)\n",
    "    # Build prediction sets for APS and RAPS with different values of alpha\n",
    "    # and compute metrics\n",
    "    for alpha in alphas:\n",
    "        # Prediction\n",
    "        y_pred_aps, set_pred_aps = imagewoof_aps.predict(X_test, alpha=alpha)\n",
    "        y_pred_raps, set_pred_raps = imagewoof_raps.predict(X_test, alpha=alpha)\n",
    "\n",
    "        # Compute average coverage and size\n",
    "        coverage_aps = classification_mean_coverage(y_test, set_pred_aps)\n",
    "        coverage_raps = classification_mean_coverage(y_test, set_pred_raps)\n",
    "        mean_size_aps = classification_mean_size(set_pred_aps)\n",
    "        mean_size_raps = classification_mean_size(set_pred_raps)\n",
    "\n",
    "        # Append iteration results\n",
    "        results.append([\"APS Imagewoof\", alpha, coverage_aps, mean_size_aps])\n",
    "        results.append([\"RAPS Imagewoof\", alpha, coverage_raps, mean_size_raps])\n",
    "    break\n",
    "\n",
    "with open(\"results/results_classification_imagewoof.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(results, output_file)\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from deel.puncc.metrics import (\n",
    "    classification_mean_coverage,\n",
    "    classification_mean_size,\n",
    ")\n",
    "\n",
    "# Load results on Imagenette\n",
    "with open(\"results/results_classification_imagenette.pickle\", \"rb\") as input_file:\n",
    "    imagenette_results = pickle.load(input_file)\n",
    "\n",
    "results_df_imagenette = pd.DataFrame(\n",
    "    imagenette_results, columns=[\"method\", \"alpha\", \"coverage\", \"size\"]\n",
    ")\n",
    "\n",
    "# Load results on Imagewoof\n",
    "with open(\"results/results_classification_imagewoof.pickle\", \"rb\") as input_file:\n",
    "    imagewoof_results = pickle.load(input_file)\n",
    "\n",
    "results_df_imagewoof = pd.DataFrame(\n",
    "    imagewoof_results, columns=[\"method\", \"alpha\", \"coverage\", \"size\"]\n",
    ")\n",
    "\n",
    "# Validity\n",
    "results_df_imagenette[\"ace\"] = abs(\n",
    "    1 - results_df_imagenette[\"alpha\"] - results_df_imagenette[\"coverage\"]\n",
    ")\n",
    "results_df_imagewoof[\"ace\"] = abs(\n",
    "    1 - results_df_imagewoof[\"alpha\"] - results_df_imagewoof[\"coverage\"]\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>size</th>\n",
       "      <th>ace</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>method</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>APS Imagewoof</th>\n",
       "      <td>40.631727</td>\n",
       "      <td>0.025499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RAPS Imagewoof</th>\n",
       "      <td>21.207737</td>\n",
       "      <td>0.025445</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     size       ace\n",
       "method                             \n",
       "APS Imagewoof   40.631727  0.025499\n",
       "RAPS Imagewoof  21.207737  0.025445"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Compute average size and coverage gap\n",
    "results_df_imagewoof.groupby([\"method\"])[[\"size\", \"ace\"]].mean()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import scienceplots\n",
    "plt.style.use(['science'])\n",
    "\n",
    "plt.rcParams['font.size'] = 14\n",
    "\n",
    "alphas = np.arange(.1, 1, .1)\n",
    "\n",
    "markers_imagenette = [\"-o\", \"-x\"]\n",
    "\n",
    "markers_imagewoof = [\"-d\", \"-p\"]\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(6,6))\n",
    "\n",
    "labels_imagenette = [\"APS (Imagenette)\", \"RAPS (Imagenette)\"]\n",
    "labels_imagewoof = [\"APS (Imagewoof)\", \"RAPS (Imagewoof)\"]\n",
    "\n",
    "\n",
    "for i, method in enumerate(results_df_imagenette[\"method\"].unique()):\n",
    "    filter = results_df_imagenette[\"method\"] == method\n",
    "    ax.plot(1-results_df_imagenette[filter][\"alpha\"], results_df_imagenette[filter][\"coverage\"], markers_imagenette[i], label=labels_imagenette[i], markersize=5)\n",
    "\n",
    "for i, method in enumerate(results_df_imagewoof[\"method\"].unique()):\n",
    "    filter = results_df_imagewoof[\"method\"] == method\n",
    "    ax.plot(1-results_df_imagewoof[filter][\"alpha\"], results_df_imagewoof[filter][\"coverage\"], markers_imagewoof[i], label=labels_imagewoof[i], markersize=5)\n",
    "\n",
    "ax.plot(np.arange(0,1.1,.1), np.arange(0,1.1,.1), \"--\", c=\"k\", label=\"Perfectly calibrated\", markersize=5)\n",
    "\n",
    "ax.set_xlabel(r\"Target Coverage ($1-\\alpha$)\")\n",
    "ax.set_ylabel(r\"Average Empirical Coverage\")\n",
    "\n",
    "ax.set_xlim([0.07,.93])\n",
    "ax.set_ylim([0.07,.93])\n",
    "\n",
    "ax.grid(True)\n",
    "\n",
    "plt.legend()\n",
    "\n",
    "plt.savefig(\"classification_coverage.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(6, 6))\n",
    "\n",
    "for i, method in enumerate(results_df_imagenette[\"method\"].unique()):\n",
    "    filter = results_df_imagenette[\"method\"] == method\n",
    "    ax.plot(\n",
    "        1 - results_df_imagenette[filter][\"alpha\"],\n",
    "        results_df_imagenette[filter][\"size\"],\n",
    "        markers_imagenette[i],\n",
    "        label=labels_imagenette[i],\n",
    "        markersize=5,\n",
    "    )\n",
    "\n",
    "for i, method in enumerate(results_df_imagewoof[\"method\"].unique()):\n",
    "    filter = results_df_imagewoof[\"method\"] == method\n",
    "    ax.plot(\n",
    "        1 - results_df_imagewoof[filter][\"alpha\"],\n",
    "        results_df_imagewoof[filter][\"size\"],\n",
    "        markers_imagewoof[i],\n",
    "        label=labels_imagewoof[i],\n",
    "        markersize=5,\n",
    "    )\n",
    "\n",
    "ax.set_xlabel(r\"Target Coverage ($1-\\alpha$)\")\n",
    "ax.set_ylabel(r\"Average Set Size\")\n",
    "\n",
    "ax.set_xlim([0.07, 0.93])\n",
    "\n",
    "ax.grid(True)\n",
    "\n",
    "plt.legend()\n",
    "\n",
    "plt.savefig(\"classification_size.pdf\")\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "puncc-dev-env",
   "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": "3.9.7"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "5d563b04d2d4b2203453d9712764b1a393ce373fc5e580bc9bc85f30c8eeb84a"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}