|
| 1 | +{ |
| 2 | + "cells": [ |
| 3 | + { |
| 4 | + "cell_type": "markdown", |
| 5 | + "metadata": {}, |
| 6 | + "source": [ |
| 7 | + "# Build Dataset\n", |
| 8 | + "Split our raw test data into train/dev/test folders.\n", |
| 9 | + "\n", |
| 10 | + "We do a split of `80%/10%/10%`" |
| 11 | + ] |
| 12 | + }, |
| 13 | + { |
| 14 | + "cell_type": "code", |
| 15 | + "execution_count": 69, |
| 16 | + "metadata": {}, |
| 17 | + "outputs": [], |
| 18 | + "source": [ |
| 19 | + "import os\n", |
| 20 | + "import random\n", |
| 21 | + "from tqdm import tqdm\n", |
| 22 | + "from shutil import copyfile" |
| 23 | + ] |
| 24 | + }, |
| 25 | + { |
| 26 | + "cell_type": "code", |
| 27 | + "execution_count": 70, |
| 28 | + "metadata": {}, |
| 29 | + "outputs": [], |
| 30 | + "source": [ |
| 31 | + "raw_data_dir = './raw_data/'\n", |
| 32 | + "assert os.path.isdir(data_dir)\n", |
| 33 | + "\n", |
| 34 | + "output_dir = './processed_data/'" |
| 35 | + ] |
| 36 | + }, |
| 37 | + { |
| 38 | + "cell_type": "code", |
| 39 | + "execution_count": 71, |
| 40 | + "metadata": {}, |
| 41 | + "outputs": [], |
| 42 | + "source": [ |
| 43 | + "filenames = os.listdir(raw_data_dir)\n", |
| 44 | + "\n", |
| 45 | + "# Transform filenames into pairs (.gui, .png)\n", |
| 46 | + "filenames = [(f[:-3] + 'gui', f[:-3] + 'png') for f in filenames if f.endswith('.gui')]\n", |
| 47 | + "\n", |
| 48 | + "random.seed(12345)\n", |
| 49 | + "filenames.sort()\n", |
| 50 | + "random.shuffle(filenames)\n", |
| 51 | + "\n", |
| 52 | + "split_1 = int(0.8 * len(filenames))\n", |
| 53 | + "split_2 = int(0.9 * len(filenames))\n", |
| 54 | + "\n", |
| 55 | + "filenames = {\n", |
| 56 | + " 'train': filenames[:split_1],\n", |
| 57 | + " 'dev': filenames[split_1:split_2],\n", |
| 58 | + " 'test': filenames[split_2:]\n", |
| 59 | + "}" |
| 60 | + ] |
| 61 | + }, |
| 62 | + { |
| 63 | + "cell_type": "code", |
| 64 | + "execution_count": 72, |
| 65 | + "metadata": {}, |
| 66 | + "outputs": [ |
| 67 | + { |
| 68 | + "name": "stdout", |
| 69 | + "output_type": "stream", |
| 70 | + "text": [ |
| 71 | + "Warning: output dir ./processed_data/ already exists.\n" |
| 72 | + ] |
| 73 | + } |
| 74 | + ], |
| 75 | + "source": [ |
| 76 | + "if not os.path.exists(output_dir):\n", |
| 77 | + " os.mkdir(output_dir)\n", |
| 78 | + "else:\n", |
| 79 | + " print('Warning: output dir {} already exists.'.format(output_dir))" |
| 80 | + ] |
| 81 | + }, |
| 82 | + { |
| 83 | + "cell_type": "code", |
| 84 | + "execution_count": 73, |
| 85 | + "metadata": {}, |
| 86 | + "outputs": [ |
| 87 | + { |
| 88 | + "name": "stderr", |
| 89 | + "output_type": "stream", |
| 90 | + "text": [ |
| 91 | + " 20%|█▉ | 275/1400 [00:00<00:00, 2741.09it/s]" |
| 92 | + ] |
| 93 | + }, |
| 94 | + { |
| 95 | + "name": "stdout", |
| 96 | + "output_type": "stream", |
| 97 | + "text": [ |
| 98 | + "Warning: output dir ./processed_data/data_train already exists.\n", |
| 99 | + "Processing train data, saving to ./processed_data/data_train.\n" |
| 100 | + ] |
| 101 | + }, |
| 102 | + { |
| 103 | + "name": "stderr", |
| 104 | + "output_type": "stream", |
| 105 | + "text": [ |
| 106 | + "100%|██████████| 1400/1400 [00:00<00:00, 2771.02it/s]\n", |
| 107 | + "100%|██████████| 175/175 [00:00<00:00, 2749.07it/s]\n", |
| 108 | + "100%|██████████| 175/175 [00:00<00:00, 2787.35it/s]" |
| 109 | + ] |
| 110 | + }, |
| 111 | + { |
| 112 | + "name": "stdout", |
| 113 | + "output_type": "stream", |
| 114 | + "text": [ |
| 115 | + "Warning: output dir ./processed_data/data_dev already exists.\n", |
| 116 | + "Processing dev data, saving to ./processed_data/data_dev.\n", |
| 117 | + "Warning: output dir ./processed_data/data_test already exists.\n", |
| 118 | + "Processing test data, saving to ./processed_data/data_test.\n" |
| 119 | + ] |
| 120 | + }, |
| 121 | + { |
| 122 | + "name": "stderr", |
| 123 | + "output_type": "stream", |
| 124 | + "text": [ |
| 125 | + "\n" |
| 126 | + ] |
| 127 | + } |
| 128 | + ], |
| 129 | + "source": [ |
| 130 | + "for split in ['train', 'dev', 'test']:\n", |
| 131 | + " output_dir_split = os.path.join(output_dir, 'data_{}'.format(split))\n", |
| 132 | + " \n", |
| 133 | + " if not os.path.exists(output_dir_split):\n", |
| 134 | + " os.mkdir(output_dir_split)\n", |
| 135 | + " else:\n", |
| 136 | + " print('Warning: output dir {} already exists.'.format(output_dir_split))\n", |
| 137 | + " \n", |
| 138 | + " print('Processing {} data, saving to {}.'.format(split, output_dir_split))\n", |
| 139 | + " \n", |
| 140 | + " for (gui, png) in tqdm(filenames[split]):\n", |
| 141 | + " src_path_gui = os.path.join(raw_data_dir, gui)\n", |
| 142 | + " output_path_gui = os.path.join(output_dir_split, gui)\n", |
| 143 | + " src_path_png = os.path.join(raw_data_dir, png)\n", |
| 144 | + " output_path_png = os.path.join(output_dir_split, png)\n", |
| 145 | + " \n", |
| 146 | + " copyfile(src_path_gui, output_path_gui)\n", |
| 147 | + " copyfile(src_path_png, output_path_png)" |
| 148 | + ] |
| 149 | + }, |
| 150 | + { |
| 151 | + "cell_type": "code", |
| 152 | + "execution_count": null, |
| 153 | + "metadata": {}, |
| 154 | + "outputs": [], |
| 155 | + "source": [] |
| 156 | + } |
| 157 | + ], |
| 158 | + "metadata": { |
| 159 | + "kernelspec": { |
| 160 | + "display_name": "Environment (conda_pytorch_p36)", |
| 161 | + "language": "python", |
| 162 | + "name": "conda_pytorch_p36" |
| 163 | + }, |
| 164 | + "language_info": { |
| 165 | + "codemirror_mode": { |
| 166 | + "name": "ipython", |
| 167 | + "version": 3 |
| 168 | + }, |
| 169 | + "file_extension": ".py", |
| 170 | + "mimetype": "text/x-python", |
| 171 | + "name": "python", |
| 172 | + "nbconvert_exporter": "python", |
| 173 | + "pygments_lexer": "ipython3", |
| 174 | + "version": "3.6.4" |
| 175 | + } |
| 176 | + }, |
| 177 | + "nbformat": 4, |
| 178 | + "nbformat_minor": 2 |
| 179 | +} |
0 commit comments