upgini 1.2.60a3792.dev1__py3-none-any.whl → 1.2.60a3792.dev2__py3-none-any.whl

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.

Potentially problematic release.


This version of upgini might be problematic. Click here for more details.

upgini/__about__.py CHANGED
@@ -1 +1 @@
1
- __version__ = "1.2.60a3792.dev1"
1
+ __version__ = "1.2.60a3792.dev2"
upgini/dataset.py CHANGED
@@ -40,7 +40,6 @@ from upgini.utils.email_utils import EmailSearchKeyConverter
40
40
  from upgini.utils.target_utils import (
41
41
  balance_undersample,
42
42
  balance_undersample_forced,
43
- balance_undersample_time_series,
44
43
  balance_undersample_time_series_trunc,
45
44
  )
46
45
 
@@ -304,12 +303,12 @@ class Dataset: # (pd.DataFrame):
304
303
  )
305
304
 
306
305
  # Resample over fit threshold
307
- if not self.imbalanced and EVAL_SET_INDEX in self.data.columns:
308
- sample_threshold = self.FIT_SAMPLE_WITH_EVAL_SET_THRESHOLD
309
- sample_rows = self.FIT_SAMPLE_WITH_EVAL_SET_ROWS
310
- elif self.cv_type is not None and self.cv_type.is_time_series():
306
+ if self.cv_type is not None and self.cv_type.is_time_series():
311
307
  sample_threshold = self.FIT_SAMPLE_THRESHOLD_TS
312
308
  sample_rows = self.FIT_SAMPLE_ROWS_TS
309
+ elif not self.imbalanced and EVAL_SET_INDEX in self.data.columns:
310
+ sample_threshold = self.FIT_SAMPLE_WITH_EVAL_SET_THRESHOLD
311
+ sample_rows = self.FIT_SAMPLE_WITH_EVAL_SET_ROWS
313
312
  else:
314
313
  sample_threshold = self.FIT_SAMPLE_THRESHOLD
315
314
  sample_rows = self.FIT_SAMPLE_ROWS
@@ -313,7 +313,7 @@ def balance_undersample_time_series_trunc(
313
313
 
314
314
  if time_unit < time_unit_threshold:
315
315
  for trunc_length in highfreq_trunc_lengths:
316
- sampled_df = trunc_datetime(dates_df, id_columns, date_column, trunc_length)
316
+ sampled_df = trunc_datetime(dates_df, id_columns, date_column, trunc_length, logger=logger)
317
317
  if len(sampled_df) <= sample_size:
318
318
  break
319
319
  if len(sampled_df) > sample_size:
@@ -322,7 +322,7 @@ def balance_undersample_time_series_trunc(
322
322
  )
323
323
  else:
324
324
  for trunc_length in lowfreq_trunc_lengths:
325
- sampled_df = trunc_datetime(dates_df, id_columns, date_column, trunc_length)
325
+ sampled_df = trunc_datetime(dates_df, id_columns, date_column, trunc_length, logger=logger)
326
326
  if len(sampled_df) <= sample_size:
327
327
  break
328
328
  if len(sampled_df) > sample_size:
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.3
2
2
  Name: upgini
3
- Version: 1.2.60a3792.dev1
3
+ Version: 1.2.60a3792.dev2
4
4
  Summary: Intelligent data search & enrichment for Machine Learning
5
5
  Project-URL: Bug Reports, https://github.com/upgini/upgini/issues
6
6
  Project-URL: Homepage, https://upgini.com/
@@ -1,7 +1,7 @@
1
- upgini/__about__.py,sha256=yYjoHiqKj96yFzYqXlsnJPzF_FcgZvyGwKBQjTVsNi4,33
1
+ upgini/__about__.py,sha256=8VHknGDFZaUmYuHXftocfi6wQkP1htM0HF9T5bSV43M,33
2
2
  upgini/__init__.py,sha256=LXSfTNU0HnlOkE69VCxkgIKDhWP-JFo_eBQ71OxTr5Y,261
3
3
  upgini/ads.py,sha256=nvuRxRx5MHDMgPr9SiU-fsqRdFaBv8p4_v1oqiysKpc,2714
4
- upgini/dataset.py,sha256=iSZX4KiDJlJFukNnAzBgkuT3UqbS-pyOyJlVXwTyaU0,34993
4
+ upgini/dataset.py,sha256=OGjpeFHbj3lWiZTOHTpWEoMMDmFY1FlNC44FKktoZvU,34956
5
5
  upgini/errors.py,sha256=2b_Wbo0OYhLUbrZqdLIx5jBnAsiD1Mcenh-VjR4HCTw,950
6
6
  upgini/features_enricher.py,sha256=IXU6ahvQqMGLdZsrHCjOGEia1pBAgixfld3pNVPcGEM,202468
7
7
  upgini/http.py,sha256=ud0Cp7h0jNeHuuZGpU_1dAAEiabGoJjGxc1X5oeBQr4,43496
@@ -56,11 +56,11 @@ upgini/utils/phone_utils.py,sha256=IrbztLuOJBiePqqxllfABWfYlfAjYevPhXKipl95wUI,1
56
56
  upgini/utils/postal_code_utils.py,sha256=5M0sUqH2DAr33kARWCTXR-ACyzWbjDq_-0mmEml6ZcU,1716
57
57
  upgini/utils/progress_bar.py,sha256=N-Sfdah2Hg8lXP_fV9EfUTXz_PyRt4lo9fAHoUDOoLc,1550
58
58
  upgini/utils/sklearn_ext.py,sha256=13jQS_k7v0aUtudXV6nGUEWjttPQzAW9AFYL5wgEz9k,44511
59
- upgini/utils/target_utils.py,sha256=a7Ck7WgQeUhDrnluOdFXvOdX6zDL-4Wiqt_f4jZxHag,16543
59
+ upgini/utils/target_utils.py,sha256=gGIO40NmLdm1DWaZNPWPoCIPlLNIDZWFGjoPuhOlzAU,16573
60
60
  upgini/utils/track_info.py,sha256=G5Lu1xxakg2_TQjKZk4b5SvrHsATTXNVV3NbvWtT8k8,5663
61
61
  upgini/utils/ts_utils.py,sha256=_YbNVE144vtEPlvLpvPGguDNzrnUM9IIjdX2VQz4T7E,1671
62
62
  upgini/utils/warning_counter.py,sha256=-GRY8EUggEBKODPSuXAkHn9KnEQwAORC0mmz_tim-PM,254
63
- upgini-1.2.60a3792.dev1.dist-info/METADATA,sha256=4k4LdGfGvuhNHhpT83pomgnfvZr8x2fKQDQbFCEAyPA,49065
64
- upgini-1.2.60a3792.dev1.dist-info/WHEEL,sha256=zEMcRr9Kr03x1ozGwg5v9NQBKn3kndp6LSoSlVg-jhU,87
65
- upgini-1.2.60a3792.dev1.dist-info/licenses/LICENSE,sha256=5RRzgvdJUu3BUDfv4bzVU6FqKgwHlIay63pPCSmSgzw,1514
66
- upgini-1.2.60a3792.dev1.dist-info/RECORD,,
63
+ upgini-1.2.60a3792.dev2.dist-info/METADATA,sha256=9ViiZMTzdvTgcXn1tdN-TUgwOW7ovx7JoOQk_VZfMf0,49065
64
+ upgini-1.2.60a3792.dev2.dist-info/WHEEL,sha256=zEMcRr9Kr03x1ozGwg5v9NQBKn3kndp6LSoSlVg-jhU,87
65
+ upgini-1.2.60a3792.dev2.dist-info/licenses/LICENSE,sha256=5RRzgvdJUu3BUDfv4bzVU6FqKgwHlIay63pPCSmSgzw,1514
66
+ upgini-1.2.60a3792.dev2.dist-info/RECORD,,