CSE515_MWDB_Project/Phase 3/task_4.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "from utils import *\n",
    "warnings.filterwarnings('ignore')\n",
    "%matplotlib inline\n",
    "from sklearn.decomposition import PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "fd_collection = getCollection(\"team_5_mwdb_phase_2\", \"fd_collection\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "selected_feature_model = \"fc_fd\"\n",
    "\n",
    "t = int(input(\"Enter value of t: \"))\n",
    "if t < 1:\n",
    "    raise ValueError(\"t should be a positive integer\")\n",
    "\n",
    "num_layers = int(input(\"Enter the number of layers: \"))\n",
    "if num_layers < 1:\n",
    "    raise ValueError(\"num_layers should be a positive integer\")\n",
    "\n",
    "num_hashes_per_layer = int(input(\"Enter the number of hashes per layer: \"))\n",
    "if num_hashes_per_layer < 1:\n",
    "    raise ValueError(\"num_hashes_per_layer should be a positive integer\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_images = list(fd_collection.find())\n",
    "all_images = sorted(all_images, key = lambda x: x[\"image_id\"])\n",
    "\n",
    "odd_image_ids = [img[\"image_id\"] for img in all_images if img[\"image_id\"] % 2 == 0]\n",
    "\n",
    "even_image_labels = [img[\"true_label\"] for img in all_images if img[\"image_id\"] % 2 == 0]\n",
    "odd_image_labels = [img[\"true_label\"] for img in all_images if img[\"image_id\"] % 2 != 0]\n",
    "\n",
    "feature_vectors = [np.array(img[selected_feature_model]).flatten() for img in all_images]\n",
    "\n",
    "total_len = len(feature_vectors)\n",
    "even_feature_vectors = []\n",
    "odd_feature_vectors = []\n",
    "\n",
    "for i in range(total_len):\n",
    "  if i % 2 == 0:\n",
    "    even_feature_vectors.append(feature_vectors[i])\n",
    "  else:\n",
    "    odd_feature_vectors.append(feature_vectors[i])\n",
    "\n",
    "\n",
    "even_feature_vectors = np.array(even_feature_vectors)\n",
    "odd_feature_vectors = np.array(odd_feature_vectors)\n",
    "\n",
    "odd_len = odd_feature_vectors.shape[0]\n",
    "even_len = even_feature_vectors.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_vectors = even_feature_vectors.shape[0]\n",
    "vector_dimension = even_feature_vectors.shape[1]\n",
    "\n",
    "# lsh = LSHIndex(num_layers, num_hashes_per_layer, vector_dimension)\n",
    "\n",
    "# for index, vector in enumerate(even_feature_vectors):\n",
    "#     lsh.add_vector(vector.tolist(), index * 2)\n",
    "\n",
    "lsh = LSH(even_feature_vectors, num_layers, num_hashes_per_layer)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of unique images considered: 1842\n",
      "Overall number of images considered: 1842 \n",
      "\n",
      "Input image ID: 8675\n",
      "Input image label: 100\n",
      "10 images\n",
      "Label: 61, ID: 6328\n",
      "Label: 80, ID: 7314\n",
      "Label: 100, ID: 8630\n",
      "Label: 80, ID: 7306\n",
      "Label: 99, ID: 8600\n",
      "Label: 63, ID: 6482\n",
      "Label: 100, ID: 8624\n",
      "Label: 99, ID: 8594\n",
      "Label: 100, ID: 8626\n",
      "Label: 6, ID: 2724\n"
     ]
    }
   ],
   "source": [
    "query_image_id = int(input(\"Enter value of query_image_id: \"))\n",
    "if query_image_id < 1:\n",
    "    raise ValueError(\"query_image_id should be a positive integer\")\n",
    "\n",
    "query_vector = np.array(odd_feature_vectors[(query_image_id // 2)])\n",
    "\n",
    "similar_images = lsh.find_similar(query_vector, t)\n",
    "unique_images_count = lsh.get_unique_images_considered_count()\n",
    "overall_images_count = lsh.get_overall_images_considered_count()\n",
    "\n",
    "\n",
    "print(\"Number of unique images considered:\", unique_images_count)\n",
    "print(\"Overall number of images considered:\", overall_images_count, \"\\n\")\n",
    "\n",
    "print(f\"Input image ID: {query_image_id}\")\n",
    "print(f\"Input image label: {odd_image_labels[query_image_id // 2]}\")\n",
    "print(f\"{len(similar_images)} images\")\n",
    "for sim in similar_images:\n",
    "    print(f\"Label: {even_image_labels[sim]}, ID: {sim * 2}\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}