{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d6299e4c-f1ba-4be4-ac89-63c05287387c",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/admin/home-devforfu/realfake\n"
     ]
    }
   ],
   "source": [
    "%cd .."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "8729a815-81b7-4667-bb97-c85456ad86e8",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "from pathlib import Path\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "from realfake.utils import list_files, write_jsonl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "82fa1fee-3ea0-415e-8a6a-8622bc55fabd",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "root = Path(\"/fsx/home-devforfu/data\")\n",
    "laion = list_files(root/\"laionimages\", [\"jpg\"])\n",
    "movie = list_files(root/\"shotcafe\", [\"jpg\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "58d79304-bdf0-445d-b5dd-46bd2a7bf1ec",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5133, 7526)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(laion), len(movie)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "f78ced08-3a5a-48f8-a635-6d93e82856c1",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "n_test = 0.1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "7a89dac3-cc8f-46cf-b60a-9118121ff69a",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "tst = set(train_test_split(np.arange(len(laion)), test_size=n_test)[1])\n",
    "metadata = [{\"path\": str(fn), \"label\": \"real\", \"valid\": i in tst} for i, fn in enumerate(laion)]\n",
    "tst = set(train_test_split(np.arange(len(movie)), test_size=n_test)[1])\n",
    "metadata += [{\"path\": str(fn), \"label\": \"fake\", \"valid\": i in tst} for i, fn in enumerate(movie)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "b0da02f5-eec6-4114-81fa-901aef933d72",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "False    11392\n",
      "True      1267\n",
      "Name: valid, dtype: int64\n",
      "fake    7526\n",
      "real    5133\n",
      "Name: label, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "df = pd.DataFrame(metadata)\n",
    "print(df.valid.value_counts())\n",
    "print(df.label.value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "bdad7b5a-17de-424b-8bfb-d13fe197552b",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "write_jsonl(\"metadata/movies.jsonl\", metadata)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "428dda09-77b8-4f6c-b14e-059e9d281f9a",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}