Build Dataset.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Build Dataset\n",
    "Split our raw test data into train/dev/test folders.\n",
    "\n",
    "We do a split of `80%/10%/10%`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import random\n",
    "from tqdm import tqdm\n",
    "from shutil import copyfile"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "raw_data_dir = './raw_data/'\n",
    "assert os.path.isdir(data_dir)\n",
    "\n",
    "output_dir = './processed_data/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "filenames = os.listdir(raw_data_dir)\n",
    "\n",
    "# Transform filenames into pairs (.gui, .png)\n",
    "filenames = [(f[:-3] + 'gui', f[:-3] + 'png') for f in filenames if f.endswith('.gui')]\n",
    "\n",
    "random.seed(12345)\n",
    "filenames.sort()\n",
    "random.shuffle(filenames)\n",
    "\n",
    "split_1 = int(0.8 * len(filenames))\n",
    "split_2 = int(0.9 * len(filenames))\n",
    "\n",
    "filenames = {\n",
    "    'train': filenames[:split_1],\n",
    "    'dev': filenames[split_1:split_2],\n",
    "    'test': filenames[split_2:]\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Warning: output dir ./processed_data/ already exists.\n"
     ]
    }
   ],
   "source": [
    "if not os.path.exists(output_dir):\n",
    "    os.mkdir(output_dir)\n",
    "else:\n",
    "    print('Warning: output dir {} already exists.'.format(output_dir))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 20%|█▉        | 275/1400 [00:00<00:00, 2741.09it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Warning: output dir ./processed_data/data_train already exists.\n",
      "Processing train data, saving to ./processed_data/data_train.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1400/1400 [00:00<00:00, 2771.02it/s]\n",
      "100%|██████████| 175/175 [00:00<00:00, 2749.07it/s]\n",
      "100%|██████████| 175/175 [00:00<00:00, 2787.35it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Warning: output dir ./processed_data/data_dev already exists.\n",
      "Processing dev data, saving to ./processed_data/data_dev.\n",
      "Warning: output dir ./processed_data/data_test already exists.\n",
      "Processing test data, saving to ./processed_data/data_test.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "for split in ['train', 'dev', 'test']:\n",
    "    output_dir_split = os.path.join(output_dir, 'data_{}'.format(split))\n",
    "    \n",
    "    if not os.path.exists(output_dir_split):\n",
    "        os.mkdir(output_dir_split)\n",
    "    else:\n",
    "        print('Warning: output dir {} already exists.'.format(output_dir_split))\n",
    "        \n",
    "    print('Processing {} data, saving to {}.'.format(split, output_dir_split))\n",
    "    \n",
    "    for (gui, png) in tqdm(filenames[split]):\n",
    "        src_path_gui = os.path.join(raw_data_dir, gui)\n",
    "        output_path_gui = os.path.join(output_dir_split, gui)\n",
    "        src_path_png = os.path.join(raw_data_dir, png)\n",
    "        output_path_png = os.path.join(output_dir_split, png)\n",
    "        \n",
    "        copyfile(src_path_gui, output_path_gui)\n",
    "        copyfile(src_path_png, output_path_png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Environment (conda_pytorch_p36)",
   "language": "python",
   "name": "conda_pytorch_p36"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}