pytae 2.2.0__tar.gz → 2.2.2__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (39) hide show
  1. {pytae-2.2.0 → pytae-2.2.2}/.ipynb_checkpoints/pyproject-checkpoint.toml +1 -1
  2. pytae-2.2.2/.ipynb_checkpoints/requirements-checkpoint.txt +2 -0
  3. {pytae-2.2.0 → pytae-2.2.2}/PKG-INFO +1 -1
  4. {pytae-2.2.0 → pytae-2.2.2}/pyproject.toml +1 -1
  5. pytae-2.2.2/requirements.txt +2 -0
  6. pytae-2.2.2/src/.ipynb_checkpoints/agg_df-checkpoint.ipynb +461 -0
  7. pytae-2.2.2/src/.ipynb_checkpoints/get_data-checkpoint.ipynb +241 -0
  8. pytae-2.2.2/src/.ipynb_checkpoints/shape-checkpoint.ipynb +224 -0
  9. pytae-2.2.2/src/agg_df.ipynb +461 -0
  10. pytae-2.2.2/src/get_data.ipynb +241 -0
  11. pytae-2.2.2/src/pytae/.ipynb_checkpoints/__init__-checkpoint.py +4 -0
  12. pytae-2.2.2/src/pytae/.ipynb_checkpoints/agg_df-checkpoint.py +53 -0
  13. pytae-2.2.2/src/pytae/.ipynb_checkpoints/extend_pandas-checkpoint.py +55 -0
  14. pytae-2.2.2/src/pytae/.ipynb_checkpoints/get_data-checkpoint.py +52 -0
  15. pytae-2.2.0/src/pytae/.ipynb_checkpoints/extend_pandas-checkpoint.py → pytae-2.2.2/src/pytae/.ipynb_checkpoints/shape-checkpoint.py +2 -111
  16. pytae-2.2.2/src/pytae/__init__.py +4 -0
  17. pytae-2.2.2/src/pytae/agg_df.py +53 -0
  18. pytae-2.2.2/src/pytae/extend_pandas.py +55 -0
  19. pytae-2.2.2/src/pytae/get_data.py +52 -0
  20. pytae-2.2.0/src/pytae/extend_pandas.py → pytae-2.2.2/src/pytae/shape.py +2 -111
  21. pytae-2.2.2/src/shape.ipynb +221 -0
  22. pytae-2.2.0/.ipynb_checkpoints/requirements-checkpoint.txt +0 -3
  23. pytae-2.2.0/requirements.txt +0 -3
  24. pytae-2.2.0/src/.ipynb_checkpoints/recon-checkpoint.ipynb +0 -855
  25. pytae-2.2.0/src/pytae/.ipynb_checkpoints/__init__-checkpoint.py +0 -3
  26. pytae-2.2.0/src/pytae/.ipynb_checkpoints/general_utils-checkpoint.py +0 -279
  27. pytae-2.2.0/src/pytae/.ipynb_checkpoints/reconciller-checkpoint.py +0 -103
  28. pytae-2.2.0/src/pytae/__init__.py +0 -3
  29. pytae-2.2.0/src/pytae/general_utils.py +0 -279
  30. pytae-2.2.0/src/pytae/reconciller.py +0 -103
  31. pytae-2.2.0/src/recon.ipynb +0 -867
  32. {pytae-2.2.0 → pytae-2.2.2}/.github/workflows/publish-to-test-pypi.yml +0 -0
  33. {pytae-2.2.0 → pytae-2.2.2}/.gitignore +0 -0
  34. {pytae-2.2.0 → pytae-2.2.2}/.ipynb_checkpoints/LICENSE-checkpoint +0 -0
  35. {pytae-2.2.0 → pytae-2.2.2}/.ipynb_checkpoints/README-checkpoint.md +0 -0
  36. {pytae-2.2.0 → pytae-2.2.2}/LICENSE +0 -0
  37. {pytae-2.2.0 → pytae-2.2.2}/README.md +0 -0
  38. {pytae-2.2.0 → pytae-2.2.2}/src/.ipynb_checkpoints/other_examples-checkpoint.ipynb +0 -0
  39. {pytae-2.2.0 → pytae-2.2.2}/src/other_examples.ipynb +0 -0
{pytae-2.2.0 → pytae-2.2.2}/.ipynb_checkpoints/pyproject-checkpoint.toml RENAMED
@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
4
4
 
5
5
  [project]
6
6
  name = "pytae"
7
- version = "2.2.0"
7
+ version = "2.2.2"
8
8
  authors = [
9
9
  { name="Madhukar Kumar Jha", email="madhukar.k.jha@gmail.com" },
10
10
  ]
pytae-2.2.2/.ipynb_checkpoints/requirements-checkpoint.txt ADDED
@@ -0,0 +1,2 @@
1
+ pandas>=1.0.0
2
+
{pytae-2.2.0 → pytae-2.2.2}/PKG-INFO RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.3
2
2
  Name: pytae
3
- Version: 2.2.0
3
+ Version: 2.2.2
4
4
  Dynamic: Requires-Dist
5
5
  Summary: A package to enhance common day to day tasks in the life of a data scientist
6
6
  Project-URL: Homepage, https://github.com/maddytae/pytae
{pytae-2.2.0 → pytae-2.2.2}/pyproject.toml RENAMED
@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
4
4
 
5
5
  [project]
6
6
  name = "pytae"
7
- version = "2.2.0"
7
+ version = "2.2.2"
8
8
  authors = [
9
9
  { name="Madhukar Kumar Jha", email="madhukar.k.jha@gmail.com" },
10
10
  ]
pytae-2.2.2/requirements.txt ADDED
@@ -0,0 +1,2 @@
1
+ pandas>=1.0.0
2
+
pytae-2.2.2/src/.ipynb_checkpoints/agg_df-checkpoint.ipynb ADDED
@@ -0,0 +1,461 @@
1
+ {
2
+ "cells": [
3
+ {
4
+ "cell_type": "code",
5
+ "execution_count": 22,
6
+ "id": "e1e35300-a92f-47dd-b448-83861e79eedf",
7
+ "metadata": {},
8
+ "outputs": [],
9
+ "source": [
10
+ "import pandas as pd\n",
11
+ "import pytae as pt\n",
12
+ "import seaborn as sns\n",
13
+ "penguins = sns.load_dataset('penguins')"
14
+ ]
15
+ },
16
+ {
17
+ "cell_type": "code",
18
+ "execution_count": 19,
19
+ "id": "c3f73408-8dc7-4ea5-899b-e3b23a692cd8",
20
+ "metadata": {},
21
+ "outputs": [],
22
+ "source": [
23
+ "df=pd.DataFrame({'id':['a','b','c','d','e','','f','f'],\n",
24
+ " 'balance':[10,20,0,21,15,10,20,25],\n",
25
+ " 'country':['sg','cn','ca','np','in','in','np','in']})"
26
+ ]
27
+ },
28
+ {
29
+ "cell_type": "code",
30
+ "execution_count": 20,
31
+ "id": "db20c1c7-c11f-431d-ae9c-d0c424f0d423",
32
+ "metadata": {},
33
+ "outputs": [
34
+ {
35
+ "data": {
36
+ "text/html": [
37
+ "<div>\n",
38
+ "<style scoped>\n",
39
+ " .dataframe tbody tr th:only-of-type {\n",
40
+ " vertical-align: middle;\n",
41
+ " }\n",
42
+ "\n",
43
+ " .dataframe tbody tr th {\n",
44
+ " vertical-align: top;\n",
45
+ " }\n",
46
+ "\n",
47
+ " .dataframe thead th {\n",
48
+ " text-align: right;\n",
49
+ " }\n",
50
+ "</style>\n",
51
+ "<table border=\"1\" class=\"dataframe\">\n",
52
+ " <thead>\n",
53
+ " <tr style=\"text-align: right;\">\n",
54
+ " <th></th>\n",
55
+ " <th>id</th>\n",
56
+ " <th>balance_sum</th>\n",
57
+ " <th>balance_min</th>\n",
58
+ " <th>balance_mean</th>\n",
59
+ " <th>balance_max</th>\n",
60
+ " <th>n</th>\n",
61
+ " </tr>\n",
62
+ " </thead>\n",
63
+ " <tbody>\n",
64
+ " <tr>\n",
65
+ " <th>0</th>\n",
66
+ " <td></td>\n",
67
+ " <td>10</td>\n",
68
+ " <td>10</td>\n",
69
+ " <td>10.0</td>\n",
70
+ " <td>10</td>\n",
71
+ " <td>1</td>\n",
72
+ " </tr>\n",
73
+ " <tr>\n",
74
+ " <th>1</th>\n",
75
+ " <td>a</td>\n",
76
+ " <td>10</td>\n",
77
+ " <td>10</td>\n",
78
+ " <td>10.0</td>\n",
79
+ " <td>10</td>\n",
80
+ " <td>1</td>\n",
81
+ " </tr>\n",
82
+ " <tr>\n",
83
+ " <th>2</th>\n",
84
+ " <td>b</td>\n",
85
+ " <td>20</td>\n",
86
+ " <td>20</td>\n",
87
+ " <td>20.0</td>\n",
88
+ " <td>20</td>\n",
89
+ " <td>1</td>\n",
90
+ " </tr>\n",
91
+ " <tr>\n",
92
+ " <th>3</th>\n",
93
+ " <td>c</td>\n",
94
+ " <td>0</td>\n",
95
+ " <td>0</td>\n",
96
+ " <td>0.0</td>\n",
97
+ " <td>0</td>\n",
98
+ " <td>1</td>\n",
99
+ " </tr>\n",
100
+ " <tr>\n",
101
+ " <th>4</th>\n",
102
+ " <td>d</td>\n",
103
+ " <td>21</td>\n",
104
+ " <td>21</td>\n",
105
+ " <td>21.0</td>\n",
106
+ " <td>21</td>\n",
107
+ " <td>1</td>\n",
108
+ " </tr>\n",
109
+ " <tr>\n",
110
+ " <th>5</th>\n",
111
+ " <td>e</td>\n",
112
+ " <td>15</td>\n",
113
+ " <td>15</td>\n",
114
+ " <td>15.0</td>\n",
115
+ " <td>15</td>\n",
116
+ " <td>1</td>\n",
117
+ " </tr>\n",
118
+ " <tr>\n",
119
+ " <th>6</th>\n",
120
+ " <td>f</td>\n",
121
+ " <td>45</td>\n",
122
+ " <td>20</td>\n",
123
+ " <td>22.5</td>\n",
124
+ " <td>25</td>\n",
125
+ " <td>2</td>\n",
126
+ " </tr>\n",
127
+ " </tbody>\n",
128
+ "</table>\n",
129
+ "</div>"
130
+ ],
131
+ "text/plain": [
132
+ " id balance_sum balance_min balance_mean balance_max n\n",
133
+ "0 10 10 10.0 10 1\n",
134
+ "1 a 10 10 10.0 10 1\n",
135
+ "2 b 20 20 20.0 20 1\n",
136
+ "3 c 0 0 0.0 0 1\n",
137
+ "4 d 21 21 21.0 21 1\n",
138
+ "5 e 15 15 15.0 15 1\n",
139
+ "6 f 45 20 22.5 25 2"
140
+ ]
141
+ },
142
+ "execution_count": 20,
143
+ "metadata": {},
144
+ "output_type": "execute_result"
145
+ }
146
+ ],
147
+ "source": [
148
+ "df[['id','balance']].agg_df(type=['sum','min','mean','min','max','n'])"
149
+ ]
150
+ },
151
+ {
152
+ "cell_type": "code",
153
+ "execution_count": 32,
154
+ "id": "66cb16ac-30b8-491e-94b4-0a2065d1e29b",
155
+ "metadata": {},
156
+ "outputs": [
157
+ {
158
+ "data": {
159
+ "text/html": [
160
+ "<div>\n",
161
+ "<style scoped>\n",
162
+ " .dataframe tbody tr th:only-of-type {\n",
163
+ " vertical-align: middle;\n",
164
+ " }\n",
165
+ "\n",
166
+ " .dataframe tbody tr th {\n",
167
+ " vertical-align: top;\n",
168
+ " }\n",
169
+ "\n",
170
+ " .dataframe thead th {\n",
171
+ " text-align: right;\n",
172
+ " }\n",
173
+ "</style>\n",
174
+ "<table border=\"1\" class=\"dataframe\">\n",
175
+ " <thead>\n",
176
+ " <tr style=\"text-align: right;\">\n",
177
+ " <th></th>\n",
178
+ " <th>id</th>\n",
179
+ " <th>balance_sum</th>\n",
180
+ " </tr>\n",
181
+ " </thead>\n",
182
+ " <tbody>\n",
183
+ " <tr>\n",
184
+ " <th>0</th>\n",
185
+ " <td></td>\n",
186
+ " <td>10</td>\n",
187
+ " </tr>\n",
188
+ " <tr>\n",
189
+ " <th>1</th>\n",
190
+ " <td>a</td>\n",
191
+ " <td>10</td>\n",
192
+ " </tr>\n",
193
+ " <tr>\n",
194
+ " <th>2</th>\n",
195
+ " <td>b</td>\n",
196
+ " <td>20</td>\n",
197
+ " </tr>\n",
198
+ " <tr>\n",
199
+ " <th>3</th>\n",
200
+ " <td>c</td>\n",
201
+ " <td>0</td>\n",
202
+ " </tr>\n",
203
+ " <tr>\n",
204
+ " <th>4</th>\n",
205
+ " <td>d</td>\n",
206
+ " <td>21</td>\n",
207
+ " </tr>\n",
208
+ " <tr>\n",
209
+ " <th>5</th>\n",
210
+ " <td>e</td>\n",
211
+ " <td>15</td>\n",
212
+ " </tr>\n",
213
+ " <tr>\n",
214
+ " <th>6</th>\n",
215
+ " <td>f</td>\n",
216
+ " <td>45</td>\n",
217
+ " </tr>\n",
218
+ " </tbody>\n",
219
+ "</table>\n",
220
+ "</div>"
221
+ ],
222
+ "text/plain": [
223
+ " id balance_sum\n",
224
+ "0 10\n",
225
+ "1 a 10\n",
226
+ "2 b 20\n",
227
+ "3 c 0\n",
228
+ "4 d 21\n",
229
+ "5 e 15\n",
230
+ "6 f 45"
231
+ ]
232
+ },
233
+ "execution_count": 32,
234
+ "metadata": {},
235
+ "output_type": "execute_result"
236
+ }
237
+ ],
238
+ "source": [
239
+ "df[['id','balance']].agg_df()\n",
240
+ "# df[['id','balance']].agg_df() by default is returning sum; \n",
241
+ "# to do: allow it to also work like this df[['id','balance']].agg_df('mean') i.e. passing string instead of a list"
242
+ ]
243
+ },
244
+ {
245
+ "cell_type": "code",
246
+ "execution_count": 23,
247
+ "id": "cbf32f4b-5f41-4ea7-bfa4-9f884d714f31",
248
+ "metadata": {},
249
+ "outputs": [
250
+ {
251
+ "data": {
252
+ "text/html": [
253
+ "<div>\n",
254
+ "<style scoped>\n",
255
+ " .dataframe tbody tr th:only-of-type {\n",
256
+ " vertical-align: middle;\n",
257
+ " }\n",
258
+ "\n",
259
+ " .dataframe tbody tr th {\n",
260
+ " vertical-align: top;\n",
261
+ " }\n",
262
+ "\n",
263
+ " .dataframe thead th {\n",
264
+ " text-align: right;\n",
265
+ " }\n",
266
+ "</style>\n",
267
+ "<table border=\"1\" class=\"dataframe\">\n",
268
+ " <thead>\n",
269
+ " <tr style=\"text-align: right;\">\n",
270
+ " <th></th>\n",
271
+ " <th>species</th>\n",
272
+ " <th>island</th>\n",
273
+ " <th>sex</th>\n",
274
+ " <th>n</th>\n",
275
+ " <th>bill_length_mm_sum</th>\n",
276
+ " <th>bill_depth_mm_sum</th>\n",
277
+ " <th>flipper_length_mm_sum</th>\n",
278
+ " <th>body_mass_g_sum</th>\n",
279
+ " </tr>\n",
280
+ " </thead>\n",
281
+ " <tbody>\n",
282
+ " <tr>\n",
283
+ " <th>0</th>\n",
284
+ " <td>Adelie</td>\n",
285
+ " <td>Biscoe</td>\n",
286
+ " <td>Female</td>\n",
287
+ " <td>22</td>\n",
288
+ " <td>821.9</td>\n",
289
+ " <td>389.5</td>\n",
290
+ " <td>4118.0</td>\n",
291
+ " <td>74125.0</td>\n",
292
+ " </tr>\n",
293
+ " <tr>\n",
294
+ " <th>1</th>\n",
295
+ " <td>Adelie</td>\n",
296
+ " <td>Biscoe</td>\n",
297
+ " <td>Male</td>\n",
298
+ " <td>22</td>\n",
299
+ " <td>893.0</td>\n",
300
+ " <td>418.8</td>\n",
301
+ " <td>4189.0</td>\n",
302
+ " <td>89100.0</td>\n",
303
+ " </tr>\n",
304
+ " <tr>\n",
305
+ " <th>2</th>\n",
306
+ " <td>Adelie</td>\n",
307
+ " <td>Dream</td>\n",
308
+ " <td>Female</td>\n",
309
+ " <td>27</td>\n",
310
+ " <td>996.6</td>\n",
311
+ " <td>475.7</td>\n",
312
+ " <td>5072.0</td>\n",
313
+ " <td>90300.0</td>\n",
314
+ " </tr>\n",
315
+ " <tr>\n",
316
+ " <th>3</th>\n",
317
+ " <td>Adelie</td>\n",
318
+ " <td>Dream</td>\n",
319
+ " <td>Male</td>\n",
320
+ " <td>28</td>\n",
321
+ " <td>1122.0</td>\n",
322
+ " <td>527.5</td>\n",
323
+ " <td>5374.0</td>\n",
324
+ " <td>113275.0</td>\n",
325
+ " </tr>\n",
326
+ " <tr>\n",
327
+ " <th>4</th>\n",
328
+ " <td>Adelie</td>\n",
329
+ " <td>Torgersen</td>\n",
330
+ " <td>Female</td>\n",
331
+ " <td>24</td>\n",
332
+ " <td>901.3</td>\n",
333
+ " <td>421.2</td>\n",
334
+ " <td>4519.0</td>\n",
335
+ " <td>81500.0</td>\n",
336
+ " </tr>\n",
337
+ " <tr>\n",
338
+ " <th>5</th>\n",
339
+ " <td>Adelie</td>\n",
340
+ " <td>Torgersen</td>\n",
341
+ " <td>Male</td>\n",
342
+ " <td>23</td>\n",
343
+ " <td>933.5</td>\n",
344
+ " <td>446.0</td>\n",
345
+ " <td>4483.0</td>\n",
346
+ " <td>92800.0</td>\n",
347
+ " </tr>\n",
348
+ " <tr>\n",
349
+ " <th>6</th>\n",
350
+ " <td>Chinstrap</td>\n",
351
+ " <td>Dream</td>\n",
352
+ " <td>Female</td>\n",
353
+ " <td>34</td>\n",
354
+ " <td>1583.5</td>\n",
355
+ " <td>598.0</td>\n",
356
+ " <td>6519.0</td>\n",
357
+ " <td>119925.0</td>\n",
358
+ " </tr>\n",
359
+ " <tr>\n",
360
+ " <th>7</th>\n",
361
+ " <td>Chinstrap</td>\n",
362
+ " <td>Dream</td>\n",
363
+ " <td>Male</td>\n",
364
+ " <td>34</td>\n",
365
+ " <td>1737.2</td>\n",
366
+ " <td>654.6</td>\n",
367
+ " <td>6797.0</td>\n",
368
+ " <td>133925.0</td>\n",
369
+ " </tr>\n",
370
+ " <tr>\n",
371
+ " <th>8</th>\n",
372
+ " <td>Gentoo</td>\n",
373
+ " <td>Biscoe</td>\n",
374
+ " <td>Female</td>\n",
375
+ " <td>58</td>\n",
376
+ " <td>2642.7</td>\n",
377
+ " <td>825.8</td>\n",
378
+ " <td>12337.0</td>\n",
379
+ " <td>271425.0</td>\n",
380
+ " </tr>\n",
381
+ " <tr>\n",
382
+ " <th>9</th>\n",
383
+ " <td>Gentoo</td>\n",
384
+ " <td>Biscoe</td>\n",
385
+ " <td>Male</td>\n",
386
+ " <td>61</td>\n",
387
+ " <td>3017.9</td>\n",
388
+ " <td>958.8</td>\n",
389
+ " <td>13514.0</td>\n",
390
+ " <td>334575.0</td>\n",
391
+ " </tr>\n",
392
+ " </tbody>\n",
393
+ "</table>\n",
394
+ "</div>"
395
+ ],
396
+ "text/plain": [
397
+ " species island sex n bill_length_mm_sum bill_depth_mm_sum \\\n",
398
+ "0 Adelie Biscoe Female 22 821.9 389.5 \n",
399
+ "1 Adelie Biscoe Male 22 893.0 418.8 \n",
400
+ "2 Adelie Dream Female 27 996.6 475.7 \n",
401
+ "3 Adelie Dream Male 28 1122.0 527.5 \n",
402
+ "4 Adelie Torgersen Female 24 901.3 421.2 \n",
403
+ "5 Adelie Torgersen Male 23 933.5 446.0 \n",
404
+ "6 Chinstrap Dream Female 34 1583.5 598.0 \n",
405
+ "7 Chinstrap Dream Male 34 1737.2 654.6 \n",
406
+ "8 Gentoo Biscoe Female 58 2642.7 825.8 \n",
407
+ "9 Gentoo Biscoe Male 61 3017.9 958.8 \n",
408
+ "\n",
409
+ " flipper_length_mm_sum body_mass_g_sum \n",
410
+ "0 4118.0 74125.0 \n",
411
+ "1 4189.0 89100.0 \n",
412
+ "2 5072.0 90300.0 \n",
413
+ "3 5374.0 113275.0 \n",
414
+ "4 4519.0 81500.0 \n",
415
+ "5 4483.0 92800.0 \n",
416
+ "6 6519.0 119925.0 \n",
417
+ "7 6797.0 133925.0 \n",
418
+ "8 12337.0 271425.0 \n",
419
+ "9 13514.0 334575.0 "
420
+ ]
421
+ },
422
+ "execution_count": 23,
423
+ "metadata": {},
424
+ "output_type": "execute_result"
425
+ }
426
+ ],
427
+ "source": [
428
+ "penguins.agg_df(type=['n','sum']) # notice n is not repeated"
429
+ ]
430
+ },
431
+ {
432
+ "cell_type": "code",
433
+ "execution_count": null,
434
+ "id": "773cca3c-564a-4c3a-bfcf-12b5ddfa6683",
435
+ "metadata": {},
436
+ "outputs": [],
437
+ "source": []
438
+ }
439
+ ],
440
+ "metadata": {
441
+ "kernelspec": {
442
+ "display_name": "py_env",
443
+ "language": "python",
444
+ "name": "py_env"
445
+ },
446
+ "language_info": {
447
+ "codemirror_mode": {
448
+ "name": "ipython",
449
+ "version": 3
450
+ },
451
+ "file_extension": ".py",
452
+ "mimetype": "text/x-python",
453
+ "name": "python",
454
+ "nbconvert_exporter": "python",
455
+ "pygments_lexer": "ipython3",
456
+ "version": "3.9.6"
457
+ }
458
+ },
459
+ "nbformat": 4,
460
+ "nbformat_minor": 5
461
+ }