Hugging Face's logo

Spaces:
ankitkupadhyay
/
fire_and_smoke
Sleeping

App Files Community
ankitkupadhyay commited on
Commit
dce0000
·
1 Parent(s): 43789f0

Delete data_exploration.ipynb

Browse files
Files changed (1) hide show
  1. data_exploration.ipynb +0 -394
data_exploration.ipynb DELETED
@@ -1,394 +0,0 @@
1
- {
2
- "cells": [
3
- {
4
- "cell_type": "code",
5
- "execution_count": 1,
6
- "metadata": {},
7
- "outputs": [],
8
- "source": [
9
- "import pandas as pd\n",
10
- "import numpy as np\n",
11
- "from sklearn.model_selection import train_test_split\n",
12
- "import matplotlib.pyplot as plt\n",
13
- "import cv2\n",
14
- "import os\n",
15
- "from PIL import Image\n",
16
- "from tqdm import tqdm\n",
17
- "import torch\n",
18
- "from torch.utils.data import Dataset, DataLoader\n",
19
- "from torchvision import transforms"
20
- ]
21
- },
22
- {
23
- "cell_type": "markdown",
24
- "metadata": {},
25
- "source": [
26
- "# creating train and test dataset"
27
- ]
28
- },
29
- {
30
- "cell_type": "code",
31
- "execution_count": 2,
32
- "metadata": {},
33
- "outputs": [],
34
- "source": [
35
- "def getData(type):\n",
36
- " df = list()\n",
37
- " directory = f'D-Fire/{type}/labels' \n",
38
- " n = len(os.listdir(directory))\n",
39
- " for filename in tqdm(os.listdir(directory)):\n",
40
- " f = os.path.join(directory, filename)\n",
41
- " # print(f)\n",
42
- "\n",
43
- " image = filename[:-3] + 'jpg'\n",
44
- " # print(image)\n",
45
- " # break\n",
46
- " img = Image.open(f'D-Fire/{type}/images/{image}')\n",
47
- " width, height = img.size\n",
48
- " # print(width, height)\n",
49
- " # plt.imshow(img)\n",
50
- " # plt.show()\n",
51
- " # break\n",
52
- " pre = [image, width, height]\n",
53
- " if os.path.getsize(f) == 0:\n",
54
- " dp = pre + [2]\n",
55
- " df.append(dp)\n",
56
- " else:\n",
57
- " with open(f) as fp:\n",
58
- " lines = fp.readlines()\n",
59
- " for line in lines:\n",
60
- " line = line.split()\n",
61
- " line = list(map(float, line))\n",
62
- " line[0] = int(line[0])\n",
63
- " # line.insert(0, image)\n",
64
- " dp = pre + line\n",
65
- " df.append(dp)\n",
66
- " fp.close()\n",
67
- " return df, n"
68
- ]
69
- },
70
- {
71
- "cell_type": "code",
72
- "execution_count": 3,
73
- "metadata": {},
74
- "outputs": [
75
- {
76
- "name": "stderr",
77
- "output_type": "stream",
78
- "text": [
79
- "100%|██████████| 17221/17221 [00:11<00:00, 1447.90it/s]\n",
80
- "100%|██████████| 4306/4306 [00:03<00:00, 1340.39it/s]\n"
81
- ]
82
- }
83
- ],
84
- "source": [
85
- "# get train and test data\n",
86
- "train, n_train = getData(\"train\")\n",
87
- "df_train = pd.DataFrame(train, columns= [\"Image\", \"Width\", \"Height\", \"Label\", \"x_min\", \"y_min\", \"x_max\", \"y_max\"])\n",
88
- "test, n_test = getData(\"test\")\n",
89
- "df_test = pd.DataFrame(test, columns= [\"Image\", \"Width\", \"Height\", \"Label\", \"x_min\", \"y_min\", \"x_max\", \"y_max\"])"
90
- ]
91
- },
92
- {
93
- "cell_type": "code",
94
- "execution_count": 4,
95
- "metadata": {},
96
- "outputs": [
97
- {
98
- "data": {
99
- "text/html": [
100
- "<div>\n",
101
- "<style scoped>\n",
102
- " .dataframe tbody tr th:only-of-type {\n",
103
- " vertical-align: middle;\n",
104
- " }\n",
105
- "\n",
106
- " .dataframe tbody tr th {\n",
107
- " vertical-align: top;\n",
108
- " }\n",
109
- "\n",
110
- " .dataframe thead th {\n",
111
- " text-align: right;\n",
112
- " }\n",
113
- "</style>\n",
114
- "<table border=\"1\" class=\"dataframe\">\n",
115
- " <thead>\n",
116
- " <tr style=\"text-align: right;\">\n",
117
- " <th></th>\n",
118
- " <th>Image</th>\n",
119
- " <th>Width</th>\n",
120
- " <th>Height</th>\n",
121
- " <th>Label</th>\n",
122
- " <th>x_min</th>\n",
123
- " <th>y_min</th>\n",
124
- " <th>x_max</th>\n",
125
- " <th>y_max</th>\n",
126
- " </tr>\n",
127
- " </thead>\n",
128
- " <tbody>\n",
129
- " <tr>\n",
130
- " <th>0</th>\n",
131
- " <td>AoF05695.jpg</td>\n",
132
- " <td>1280</td>\n",
133
- " <td>720</td>\n",
134
- " <td>0</td>\n",
135
- " <td>0.700781</td>\n",
136
- " <td>0.379167</td>\n",
137
- " <td>0.039062</td>\n",
138
- " <td>0.105556</td>\n",
139
- " </tr>\n",
140
- " <tr>\n",
141
- " <th>1</th>\n",
142
- " <td>WEB08898.jpg</td>\n",
143
- " <td>640</td>\n",
144
- " <td>360</td>\n",
145
- " <td>0</td>\n",
146
- " <td>0.477344</td>\n",
147
- " <td>0.291667</td>\n",
148
- " <td>0.264063</td>\n",
149
- " <td>0.555556</td>\n",
150
- " </tr>\n",
151
- " <tr>\n",
152
- " <th>2</th>\n",
153
- " <td>WEB01102.jpg</td>\n",
154
- " <td>640</td>\n",
155
- " <td>360</td>\n",
156
- " <td>2</td>\n",
157
- " <td>NaN</td>\n",
158
- " <td>NaN</td>\n",
159
- " <td>NaN</td>\n",
160
- " <td>NaN</td>\n",
161
- " </tr>\n",
162
- " <tr>\n",
163
- " <th>3</th>\n",
164
- " <td>WEB07573.jpg</td>\n",
165
- " <td>1100</td>\n",
166
- " <td>619</td>\n",
167
- " <td>0</td>\n",
168
- " <td>0.465000</td>\n",
169
- " <td>0.475767</td>\n",
170
- " <td>0.290000</td>\n",
171
- " <td>0.906300</td>\n",
172
- " </tr>\n",
173
- " <tr>\n",
174
- " <th>4</th>\n",
175
- " <td>WEB08640.jpg</td>\n",
176
- " <td>640</td>\n",
177
- " <td>360</td>\n",
178
- " <td>0</td>\n",
179
- " <td>0.578125</td>\n",
180
- " <td>0.506944</td>\n",
181
- " <td>0.709375</td>\n",
182
- " <td>0.936111</td>\n",
183
- " </tr>\n",
184
- " </tbody>\n",
185
- "</table>\n",
186
- "</div>"
187
- ],
188
- "text/plain": [
189
- " Image Width Height Label x_min y_min x_max y_max\n",
190
- "0 AoF05695.jpg 1280 720 0 0.700781 0.379167 0.039062 0.105556\n",
191
- "1 WEB08898.jpg 640 360 0 0.477344 0.291667 0.264063 0.555556\n",
192
- "2 WEB01102.jpg 640 360 2 NaN NaN NaN NaN\n",
193
- "3 WEB07573.jpg 1100 619 0 0.465000 0.475767 0.290000 0.906300\n",
194
- "4 WEB08640.jpg 640 360 0 0.578125 0.506944 0.709375 0.936111"
195
- ]
196
- },
197
- "execution_count": 4,
198
- "metadata": {},
199
- "output_type": "execute_result"
200
- }
201
- ],
202
- "source": [
203
- "# train sample\n",
204
- "df_train.head()"
205
- ]
206
- },
207
- {
208
- "cell_type": "markdown",
209
- "metadata": {},
210
- "source": [
211
- "# data split exploration"
212
- ]
213
- },
214
- {
215
- "cell_type": "code",
216
- "execution_count": 5,
217
- "metadata": {},
218
- "outputs": [],
219
- "source": [
220
- "group_tr = df_train.groupby(\"Label\").count().iloc[:, 0].to_numpy()\n",
221
- "group_tr_ratio = group_tr / n_train\n",
222
- "group_te = df_test.groupby(\"Label\").count().iloc[:, 0].to_numpy()\n",
223
- "group_te_ratio = group_te / n_test"
224
- ]
225
- },
226
- {
227
- "cell_type": "code",
228
- "execution_count": 6,
229
- "metadata": {},
230
- "outputs": [
231
- {
232
- "data": {
233
- "image/png": "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",
234
- "text/plain": [
235
- "<Figure size 640x480 with 1 Axes>"
236
- ]
237
- },
238
- "metadata": {},
239
- "output_type": "display_data"
240
- }
241
- ],
242
- "source": [
243
- "# statistics on data ratio split\n",
244
- "x = np.arange(3)\n",
245
- "plt.bar(x, group_tr_ratio, color ='r', width = 0.25,\n",
246
- " edgecolor ='grey', label ='Train')\n",
247
- "x = [x + 0.25 for x in x]\n",
248
- "plt.bar(x, group_te_ratio, color ='b', width = 0.25,\n",
249
- " edgecolor ='grey', label ='Test')\n",
250
- "plt.xlabel('Labels')\n",
251
- "plt.ylabel('Proprtion Ratio')\n",
252
- "plt.xticks([0.15, 1.15, 2.15], [\"Smoke\", \"Fire\", \"None\"])\n",
253
- "plt.legend()\n",
254
- "plt.show()"
255
- ]
256
- },
257
- {
258
- "cell_type": "markdown",
259
- "metadata": {},
260
- "source": [
261
- "# total count for different classes"
262
- ]
263
- },
264
- {
265
- "cell_type": "code",
266
- "execution_count": 7,
267
- "metadata": {},
268
- "outputs": [
269
- {
270
- "data": {
271
- "text/html": [
272
- "<div>\n",
273
- "<style scoped>\n",
274
- " .dataframe tbody tr th:only-of-type {\n",
275
- " vertical-align: middle;\n",
276
- " }\n",
277
- "\n",
278
- " .dataframe tbody tr th {\n",
279
- " vertical-align: top;\n",
280
- " }\n",
281
- "\n",
282
- " .dataframe thead th {\n",
283
- " text-align: right;\n",
284
- " }\n",
285
- "</style>\n",
286
- "<table border=\"1\" class=\"dataframe\">\n",
287
- " <thead>\n",
288
- " <tr style=\"text-align: right;\">\n",
289
- " <th></th>\n",
290
- " <th>Smoke</th>\n",
291
- " <th>Fire</th>\n",
292
- " <th>Neither</th>\n",
293
- " </tr>\n",
294
- " </thead>\n",
295
- " <tbody>\n",
296
- " <tr>\n",
297
- " <th>Train</th>\n",
298
- " <td>9550</td>\n",
299
- " <td>11814</td>\n",
300
- " <td>7833</td>\n",
301
- " </tr>\n",
302
- " <tr>\n",
303
- " <th>Test</th>\n",
304
- " <td>2315</td>\n",
305
- " <td>2878</td>\n",
306
- " <td>2005</td>\n",
307
- " </tr>\n",
308
- " </tbody>\n",
309
- "</table>\n",
310
- "</div>"
311
- ],
312
- "text/plain": [
313
- " Smoke Fire Neither\n",
314
- "Train 9550 11814 7833\n",
315
- "Test 2315 2878 2005"
316
- ]
317
- },
318
- "execution_count": 7,
319
- "metadata": {},
320
- "output_type": "execute_result"
321
- }
322
- ],
323
- "source": [
324
- "pd.DataFrame([group_tr, group_te], columns=[\"Smoke\", \"Fire\", \"Neither\"], index=[\"Train\", \"Test\"])"
325
- ]
326
- },
327
- {
328
- "cell_type": "code",
329
- "execution_count": 8,
330
- "metadata": {},
331
- "outputs": [
332
- {
333
- "data": {
334
- "text/plain": [
335
- "Image Label\n",
336
- "AoF00000.jpg 2 1\n",
337
- "AoF00001.jpg 1 1\n",
338
- "AoF00002.jpg 0 1\n",
339
- "AoF00003.jpg 2 1\n",
340
- "AoF00004.jpg 2 1\n",
341
- " ..\n",
342
- "WEB09440.jpg 0 2\n",
343
- "WEB09441.jpg 0 2\n",
344
- " 1 3\n",
345
- "WEB09442.jpg 0 1\n",
346
- " 1 1\n",
347
- "Name: Width, Length: 20984, dtype: int64"
348
- ]
349
- },
350
- "execution_count": 8,
351
- "metadata": {},
352
- "output_type": "execute_result"
353
- }
354
- ],
355
- "source": [
356
- "df_train.groupby([\"Image\", \"Label\"]).count()[\"Width\"]"
357
- ]
358
- },
359
- {
360
- "cell_type": "code",
361
- "execution_count": null,
362
- "metadata": {},
363
- "outputs": [],
364
- "source": [
365
- "np.random.seed(42)\n",
366
- "idx = np.random.randint\n",
367
- "smoke = df_train[\"Label\"] == 0 \n",
368
- "fire = df_train[\"Label\"] == 1\n",
369
- "neither = df_train[\"Label\"] == 2\n"
370
- ]
371
- }
372
- ],
373
- "metadata": {
374
- "kernelspec": {
375
- "display_name": "AIClass",
376
- "language": "python",
377
- "name": "python3"
378
- },
379
- "language_info": {
380
- "codemirror_mode": {
381
- "name": "ipython",
382
- "version": 3
383
- },
384
- "file_extension": ".py",
385
- "mimetype": "text/x-python",
386
- "name": "python",
387
- "nbconvert_exporter": "python",
388
- "pygments_lexer": "ipython3",
389
- "version": "3.11.4"
390
- }
391
- },
392
- "nbformat": 4,
393
- "nbformat_minor": 2
394
- }