upgini 1.1.287a3232.post1__py3-none-any.whl → 1.1.288__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__ = "1.1.287a3232-1"
1
+ __version__ = "1.1.288"
upgini/autofe/date.py CHANGED
@@ -2,7 +2,6 @@ from typing import Any, Dict, List, Optional, Union
2
2
 
3
3
  import numpy as np
4
4
  import pandas as pd
5
- import datetime
6
5
  from pandas.core.arrays.timedeltas import TimedeltaArray
7
6
  from pydantic import BaseModel, validator
8
7
 
@@ -23,9 +22,7 @@ class DateDiffMixin(BaseModel):
23
22
  return pd.to_datetime(x, unit=unit)
24
23
 
25
24
  def _convert_diff_to_unit(self, diff: Union[pd.Series, TimedeltaArray]) -> Union[pd.Series, TimedeltaArray]:
26
- if self.diff_unit == "M":
27
- raise Exception("Unsupported difference unit: Month")
28
- elif self.diff_unit == "D":
25
+ if self.diff_unit == "D":
29
26
  if isinstance(diff, pd.Series) and diff.dtype == "object":
30
27
  return diff.apply(lambda x: None if isinstance(x, float) and np.isnan(x) else x.days)
31
28
  else:
@@ -35,6 +32,8 @@ class DateDiffMixin(BaseModel):
35
32
  return (diff / 365 / 24 / 60 / 60 / 10**9).astype(int)
36
33
  else:
37
34
  return (diff / 365 / 24 / 60 / 60 / 10**9).dt.nanoseconds
35
+ else:
36
+ raise Exception(f"Unsupported difference unit: {self.diff_unit}")
38
37
 
39
38
 
40
39
  class DateDiff(PandasOperand, DateDiffMixin):
@@ -58,6 +58,7 @@ class DataSourcePublisher:
58
58
  join_date_abs_limit_days: Optional[int] = None,
59
59
  features_for_embeddings: Optional[List[str]] = DEFAULT_GENERATE_EMBEDDINGS,
60
60
  data_table_id_to_replace: Optional[str] = None,
61
+ keep_features: Optional[List[str]] = None,
61
62
  _force_generation=False,
62
63
  _silent=False,
63
64
  ) -> str:
@@ -116,6 +117,8 @@ class DataSourcePublisher:
116
117
  request["adsDefinitionIdToReplace"] = data_table_id_to_replace
117
118
  if exclude_from_autofe_generation is not None:
118
119
  request["excludeFromGeneration"] = exclude_from_autofe_generation
120
+ if keep_features is not None:
121
+ request["keepFeatures"] = keep_features
119
122
  self.logger.info(f"Start registering data table {request}")
120
123
 
121
124
  task_id = self._rest_client.register_ads(request, trace_id)
@@ -2596,9 +2596,6 @@ class FeaturesEnricher(TransformerMixin):
2596
2596
  return validated_X
2597
2597
 
2598
2598
  def _validate_y(self, X: pd.DataFrame, y) -> pd.Series:
2599
- if _num_samples(y) == 0:
2600
- raise ValidationError(self.bundle.get("y_is_empty"))
2601
-
2602
2599
  if (
2603
2600
  not isinstance(y, pd.Series)
2604
2601
  and not isinstance(y, pd.DataFrame)
@@ -2607,6 +2604,9 @@ class FeaturesEnricher(TransformerMixin):
2607
2604
  ):
2608
2605
  raise ValidationError(self.bundle.get("unsupported_y_type").format(type(y)))
2609
2606
 
2607
+ if _num_samples(y) == 0:
2608
+ raise ValidationError(self.bundle.get("y_is_empty"))
2609
+
2610
2610
  if _num_samples(X) != _num_samples(y):
2611
2611
  raise ValidationError(self.bundle.get("x_and_y_diff_size").format(_num_samples(X), _num_samples(y)))
2612
2612
 
@@ -81,7 +81,7 @@ date_and_datetime_simultanious=DATE and DATETIME search keys cannot be used simu
81
81
  email_and_hem_simultanious=EMAIL and HEM search keys cannot be used simultaneously. Choose one to keep
82
82
  postal_code_without_country=COUNTRY search key required if POSTAL_CODE is present
83
83
  multiple_search_key=Search key {} passed multiple times
84
- unregistered_only_personal_keys=Only personal search keys used. Api_key from profile.upgini.com required for EMAIL/HEM, PHONE NUMBER or IPv4 search keys\nSee docs https://github.com/upgini/upgini#-open-up-all-capabilities-of-upgini
84
+ unregistered_only_personal_keys=Only personal search keys used. Api_key from profile.upgini.com required for EMAIL/HEM, PHONE NUMBER or IPv4/IPv6 search keys\nSee docs https://github.com/upgini/upgini#-open-up-all-capabilities-of-upgini
85
85
  search_key_not_found=Column `{}` from search_keys was not found in X dataframe: {}
86
86
  numeric_search_key_not_found=Index {} in search_keys is out of bounds for {} columns of X dataframe
87
87
  unsupported_search_key_type=Unsupported type of key in search_keys: {}
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.3
2
2
  Name: upgini
3
- Version: 1.1.287a3232.post1
3
+ Version: 1.1.288
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/
@@ -1,9 +1,9 @@
1
- upgini/__about__.py,sha256=AbRcDfYtBT7CZaikr7-ihUYbwsvTkE9kN2r6HAW5iGs,31
1
+ upgini/__about__.py,sha256=dc9KjdGN41zFpIykuStCL0GQPyjThIUdODgFyKcKALg,24
2
2
  upgini/__init__.py,sha256=asENHgEVHQBIkV-e_0IhE_ZWqkCG6398U3ZLrNzAH6k,407
3
3
  upgini/ads.py,sha256=nvuRxRx5MHDMgPr9SiU-fsqRdFaBv8p4_v1oqiysKpc,2714
4
4
  upgini/dataset.py,sha256=7TLVVhGtjgx_9yaiaIUK3kZSe_R9wg5dY0d4F5qCGM4,45636
5
5
  upgini/errors.py,sha256=2b_Wbo0OYhLUbrZqdLIx5jBnAsiD1Mcenh-VjR4HCTw,950
6
- upgini/features_enricher.py,sha256=EKngUPR5dpX8bZWrwl8qW61TtzssbxGEVwVO7XM3d3o,177530
6
+ upgini/features_enricher.py,sha256=Bfqbzqj3h1Ox3s6hBA4UYXkG7hQDqGPB4JK1YXtxXxw,177530
7
7
  upgini/http.py,sha256=khrYSldpY-HbVLCcApfV1BjBFK6Uyuatb4colKybxgY,42301
8
8
  upgini/metadata.py,sha256=qDAIO7NLSSQp_XiXCv3U4XJTLO0KH3YuQ8lvCLYPqzs,9781
9
9
  upgini/metrics.py,sha256=DLvA2YLV4f7lnzBCcfZ5T4NkqAv3pbstbjTepavuT7U,30688
@@ -15,21 +15,21 @@ upgini/ads_management/ads_manager.py,sha256=igVbN2jz80Umb2BUJixmJVj-zx8unoKpecVo
15
15
  upgini/autofe/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
16
16
  upgini/autofe/all_operands.py,sha256=7UyvmmqGSqQu4kDgoFwQRKY__b9xKDk3Fpp2-H8A7AA,2399
17
17
  upgini/autofe/binary.py,sha256=441BRuqMsxlxuw4c8rMZB6h5EpRdVMk-bVa03U7T5Hg,3973
18
- upgini/autofe/date.py,sha256=xy6D4dYvf5efVOzV8xn1H6fzYLIjjTY9MGwfnGcCAzs,7310
18
+ upgini/autofe/date.py,sha256=qzk0NT332Q0vR1eRwTuNiMSrGE3ulh6Ic3QLBZqSdvw,7284
19
19
  upgini/autofe/feature.py,sha256=_V9B74B3ue7eAYXSOt9JKhVC9klkAKks22MwnBRye_w,12487
20
20
  upgini/autofe/groupby.py,sha256=4WjDzQxqpZxB79Ih4ihMMI5GDxaFqiH6ZelfV82ClT4,3091
21
21
  upgini/autofe/operand.py,sha256=JjEVT1U3kY9NDjUPMdoki7Oa8hMDG0-_h_NklVjIFyc,2882
22
22
  upgini/autofe/unary.py,sha256=v-l3aiE5hj6kurvh6adCQL8W3X9u9a7RVbS_WPR2qlw,3146
23
23
  upgini/autofe/vector.py,sha256=dLxfAstJs-gw_OQ1xxoxcM6pVzORlV0HVzdzt7cLXVQ,606
24
24
  upgini/data_source/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
25
- upgini/data_source/data_source_publisher.py,sha256=taRzyGgrPrTTSGw4Y-Ca5k4bf30aiTa68rxqT9zfqeI,16478
25
+ upgini/data_source/data_source_publisher.py,sha256=B4fJ1owDCF5ZZ0Ca9ywi_CXVt4iPvABh5BGTnXdXmHk,16635
26
26
  upgini/mdc/__init__.py,sha256=aM08nIWFc2gWdWUa3_IuEnNND0cQPkBGnYpRMnfFN8k,1019
27
27
  upgini/mdc/context.py,sha256=3u1B-jXt7tXEvNcV3qmR9SDCseudnY7KYsLclBdwVLk,1405
28
28
  upgini/normalizer/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
29
29
  upgini/normalizer/phone_normalizer.py,sha256=EzTaahk6myRv6ZXgbyVFGY4kpo_2VlQgOrm5_lfbmNI,9996
30
30
  upgini/resource_bundle/__init__.py,sha256=S5F2G47pnJd2LDpmFsjDqEwiKkP8Hm-hcseDbMka6Ko,8345
31
31
  upgini/resource_bundle/exceptions.py,sha256=5fRvx0_vWdE1-7HcSgF0tckB4A9AKyf5RiinZkInTsI,621
32
- upgini/resource_bundle/strings.properties,sha256=1O779a0-Ai0j7W-Z5AznvjuV69YkJvgGhJda-6VMLOQ,26287
32
+ upgini/resource_bundle/strings.properties,sha256=1oHurL4I83P2lXIavx9vSdKM8ZqncAPXH2IZf76bD6g,26292
33
33
  upgini/resource_bundle/strings_widget.properties,sha256=gOdqvZWntP2LCza_tyVk1_yRYcG4c04K9sQOAVhF_gw,1577
34
34
  upgini/sampler/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
35
35
  upgini/sampler/base.py,sha256=7GpjYqjOp58vYcJLiX__1R5wjUlyQbxvHJ2klFnup_M,6389
@@ -56,7 +56,7 @@ upgini/utils/sklearn_ext.py,sha256=N-eJrfAJxYpDPc85sKQyMFcIeL9Ug2lwlqDyS4jFOdE,4
56
56
  upgini/utils/target_utils.py,sha256=Y96_PJ5cC-WsEbeqg20v9uqywDQobLoTb-xoP7S3o4E,7807
57
57
  upgini/utils/track_info.py,sha256=G5Lu1xxakg2_TQjKZk4b5SvrHsATTXNVV3NbvWtT8k8,5663
58
58
  upgini/utils/warning_counter.py,sha256=dIWBB4dI5XRRJZudvIlqlIYKEiwLLPcXarsZuYRt338,227
59
- upgini-1.1.287a3232.post1.dist-info/METADATA,sha256=WYI6YVa5RX0B1MNfStn_rU3M9QTg2Z5khzwNHxBzPp4,48128
60
- upgini-1.1.287a3232.post1.dist-info/WHEEL,sha256=zEMcRr9Kr03x1ozGwg5v9NQBKn3kndp6LSoSlVg-jhU,87
61
- upgini-1.1.287a3232.post1.dist-info/licenses/LICENSE,sha256=5RRzgvdJUu3BUDfv4bzVU6FqKgwHlIay63pPCSmSgzw,1514
62
- upgini-1.1.287a3232.post1.dist-info/RECORD,,
59
+ upgini-1.1.288.dist-info/METADATA,sha256=lUjVrD2DHFZAwdi5_HpALW7wbFRkKhu3C9rqpQwr6_k,48117
60
+ upgini-1.1.288.dist-info/WHEEL,sha256=zEMcRr9Kr03x1ozGwg5v9NQBKn3kndp6LSoSlVg-jhU,87
61
+ upgini-1.1.288.dist-info/licenses/LICENSE,sha256=5RRzgvdJUu3BUDfv4bzVU6FqKgwHlIay63pPCSmSgzw,1514
62
+ upgini-1.1.288.dist-info/RECORD,,