Delete dataset_details.ipynb

dataset_details.ipynb

@@ -1,168 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "code",
-   "execution_count": 32,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "import pandas as pd\n",
-    "import os\n",
-    "from tqdm import tqdm"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 33,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Total number of papers in the dataset: 2556\n"
-     ]
-    }
-   ],
-   "source": [
-    "# Load the dataset\n",
-    "iclr2024_df = pd.read_json('/nas/pohan/datasets/AIConfVideo/dataset/pruned_paperlist_iclr2024.json')\n",
-    "iclr2024_df['venue'] = 'iclr2024'\n",
-    "\n",
-    "nips2024_df = pd.read_json('/nas/pohan/datasets/AIConfVideo/dataset/pruned_paperlist_nips2024.json')\n",
-    "nips2024_df['venue'] = 'nips2024'\n",
-    "\n",
-    "combined_df = pd.concat([iclr2024_df, nips2024_df], ignore_index=True)\n",
-    "print(f\"Total number of papers in the dataset: {len(combined_df)}\")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 36,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Validating videos: 100%|██████████| 2556/2556 [00:18<00:00, 141.50it/s]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Total number of papers in the dataset: 2287\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n"
-     ]
-    }
-   ],
-   "source": [
-    "# Filter out rows with missing required fields\n",
-    "valid_rows = combined_df.dropna(subset=['openreview_id', 'slideslive_id', 'tldr', 'primary_area'])\n",
-    "\n",
-    "final_dataset_df = pd.DataFrame()\n",
-    "\n",
-    "# Check the validity of each corresponding video with progress bar\n",
-    "for i, row in tqdm(valid_rows.iterrows(), total=len(valid_rows), desc=\"Validating videos\"):\n",
-    "    slideslive_id = row['slideslive_id']\n",
-    "    video_dir = f\"/nas/pohan/datasets/AIConfVideo/video/{slideslive_id}\"\n",
-    "    \n",
-    "    # Check if files exist and are not empty\n",
-    "    video_path = os.path.join(video_dir, 'video.mp4')\n",
-    "    slide_path = os.path.join(video_dir, 'slide_1.png')\n",
-    "    \n",
-    "    if (os.path.exists(video_path) and os.path.getsize(video_path) > 0 and\n",
-    "        os.path.exists(slide_path) and os.path.getsize(slide_path) > 0):\n",
-    "        final_dataset_df = pd.concat([final_dataset_df, pd.DataFrame([row])], ignore_index=True)\n",
-    "\n",
-    "# Display a sample of the dataset\n",
-    "final_dataset_df.head()\n",
-    "# Print basic statistics about the dataset\n",
-    "print(f\"Total number of papers in the dataset: {len(final_dataset_df)}\")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 38,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Dataset saved to /nas/pohan/datasets/AIConfVideo/learningpaper24/catalog.json\n"
-     ]
-    }
-   ],
-   "source": [
-    "# to_save_df = final_dataset_df[['openreview_id', 'slideslive_id', 'venue', 'title', 'status', 'keywords', 'tldr', 'abstract', 'primary_area', 'site']]\n",
-    "\n",
-    "# # Save to a new JSON file\n",
-    "# output_dir = '/nas/pohan/datasets/AIConfVideo/learningpaper24'\n",
-    "# os.makedirs(output_dir, exist_ok=True)\n",
-    "# output_path = os.path.join(output_dir, 'catalog.json')\n",
-    "\n",
-    "# to_save_df.to_json(output_path, orient='records', indent=4)\n",
-    "# print(f\"Dataset saved to {output_path}\")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 51,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "# catalog_df = pd.read_json(\"/nas/pohan/datasets/AIConfVideo/learningpaper24/metadata/catalog.json\")\n",
-    "# catalog_df = catalog_df[['video_file'] + list(catalog_df.columns.drop('video_file'))]\n",
-    "\n",
-    "# catalog_df.to_json(\"/nas/pohan/datasets/AIConfVideo/learningpaper24/metadata/catalog.json\", orient='records', indent=4)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 52,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "# import json\n",
-    "\n",
-    "# # Load the JSON file\n",
-    "# with open('/nas/pohan/datasets/AIConfVideo/learningpaper24/metadata/catalog.json', 'r') as f:\n",
-    "#     data = json.load(f)  # should be a list of dicts\n",
-    "\n",
-    "# # Write out JSONL\n",
-    "# with open('/nas/pohan/datasets/AIConfVideo/learningpaper24/metadata/catalog.jsonl', 'w') as f:\n",
-    "#     for entry in data:\n",
-    "#         f.write(json.dumps(entry) + '\\n')\n"
-   ]
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "tldr",
-   "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": "3.10.16"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 2
-}