upgini 1.2.99a3922.dev5__py3-none-any.whl → 1.2.101__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 +1 -1
- upgini/features_enricher.py +8 -1
- upgini/resource_bundle/strings.properties +1 -0
- {upgini-1.2.99a3922.dev5.dist-info → upgini-1.2.101.dist-info}/METADATA +1 -1
- {upgini-1.2.99a3922.dev5.dist-info → upgini-1.2.101.dist-info}/RECORD +7 -7
- {upgini-1.2.99a3922.dev5.dist-info → upgini-1.2.101.dist-info}/WHEEL +1 -1
- {upgini-1.2.99a3922.dev5.dist-info → upgini-1.2.101.dist-info}/licenses/LICENSE +0 -0
upgini/__about__.py
CHANGED
@@ -1 +1 @@
|
|
1
|
-
__version__ = "1.2.
|
1
|
+
__version__ = "1.2.101"
|
upgini/features_enricher.py
CHANGED
@@ -3708,6 +3708,11 @@ if response.status_code == 200:
|
|
3708
3708
|
|
3709
3709
|
@staticmethod
|
3710
3710
|
def _get_group_columns(df: pd.DataFrame, search_keys: Dict[str, SearchKey]) -> List[str]:
|
3711
|
+
search_key_priority = [SearchKey.PHONE, SearchKey.EMAIL, SearchKey.HEM, SearchKey.IP]
|
3712
|
+
for key_type in search_key_priority:
|
3713
|
+
if key_type in search_keys.values():
|
3714
|
+
return [col for col, t in search_keys.items() if t == key_type]
|
3715
|
+
|
3711
3716
|
return [
|
3712
3717
|
col
|
3713
3718
|
for col, t in search_keys.items()
|
@@ -4481,7 +4486,9 @@ if response.status_code == 200:
|
|
4481
4486
|
sample = df.head(100)
|
4482
4487
|
|
4483
4488
|
def check_need_detect(search_key: SearchKey):
|
4484
|
-
return not is_transform or
|
4489
|
+
return not is_transform or (
|
4490
|
+
search_key in self.fit_search_keys.values() and search_key not in search_keys.values()
|
4491
|
+
)
|
4485
4492
|
|
4486
4493
|
# if SearchKey.POSTAL_CODE not in search_keys.values() and check_need_detect(SearchKey.POSTAL_CODE):
|
4487
4494
|
if check_need_detect(SearchKey.POSTAL_CODE):
|
@@ -99,6 +99,7 @@ unsupported_date_type=Unsupported type of date column `{}`. Convert to datetime
|
|
99
99
|
invalid_postal_code=All values of POSTAL_CODE column `{}` are invalid
|
100
100
|
invalid_country=All values of COUNTRY column `{}` are invalid
|
101
101
|
invalid_ip=All values of IP column `{}` are invalid
|
102
|
+
transform_search_keys_not_match_fit_keys=Search keys used in fit call {} are different from the search keys used in transform call. Please use the same search keys as in fit call.
|
102
103
|
# X and y validation
|
103
104
|
unsupported_x_type=Unsupported type of X: {}. Use pandas.DataFrame, pandas.Series or numpy.ndarray or list
|
104
105
|
x_contains_dup_columns=X contains duplicate column names. Please rename or drop duplicates
|
@@ -1,9 +1,9 @@
|
|
1
|
-
upgini/__about__.py,sha256=
|
1
|
+
upgini/__about__.py,sha256=cYPo00CQuvsebJbJryscL0AM9uNGVyleveinwz6t0I8,24
|
2
2
|
upgini/__init__.py,sha256=LXSfTNU0HnlOkE69VCxkgIKDhWP-JFo_eBQ71OxTr5Y,261
|
3
3
|
upgini/ads.py,sha256=nvuRxRx5MHDMgPr9SiU-fsqRdFaBv8p4_v1oqiysKpc,2714
|
4
4
|
upgini/dataset.py,sha256=e6JDYTZ2AwC5aF-dqclKZKkiKrHo2f6cFmMQO2ZZmjM,32724
|
5
5
|
upgini/errors.py,sha256=2b_Wbo0OYhLUbrZqdLIx5jBnAsiD1Mcenh-VjR4HCTw,950
|
6
|
-
upgini/features_enricher.py,sha256=
|
6
|
+
upgini/features_enricher.py,sha256=qcS6QzfAZL60bXoupXRQVWyeOodFV-hgyajvsDHIGG8,220298
|
7
7
|
upgini/http.py,sha256=DNcoS7qdxG0mOJn6I8r6O5I6XdIJTdzDzW3hkz3NgG4,45443
|
8
8
|
upgini/metadata.py,sha256=vsbbHyPCP3Rs8WkeDgQg99uAA_zmsbDStAT-NwDYhO4,12455
|
9
9
|
upgini/metrics.py,sha256=gXr2aiw5j9QBWBo1hZp40Is679hef5q8MrT6LJfjsBk,45661
|
@@ -38,7 +38,7 @@ upgini/normalizer/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU
|
|
38
38
|
upgini/normalizer/normalize_utils.py,sha256=mDh2mBW3aQMB4EFP2aHbf2dGMVkOcWnp4sKKvKDBh8w,8511
|
39
39
|
upgini/resource_bundle/__init__.py,sha256=S5F2G47pnJd2LDpmFsjDqEwiKkP8Hm-hcseDbMka6Ko,8345
|
40
40
|
upgini/resource_bundle/exceptions.py,sha256=5fRvx0_vWdE1-7HcSgF0tckB4A9AKyf5RiinZkInTsI,621
|
41
|
-
upgini/resource_bundle/strings.properties,sha256=
|
41
|
+
upgini/resource_bundle/strings.properties,sha256=NyxRwzehkrL5LMoVyjkhN811MvalepavNfjlC9ubE0Q,28677
|
42
42
|
upgini/resource_bundle/strings_widget.properties,sha256=gOdqvZWntP2LCza_tyVk1_yRYcG4c04K9sQOAVhF_gw,1577
|
43
43
|
upgini/sampler/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
|
44
44
|
upgini/sampler/base.py,sha256=7GpjYqjOp58vYcJLiX__1R5wjUlyQbxvHJ2klFnup_M,6389
|
@@ -71,7 +71,7 @@ upgini/utils/target_utils.py,sha256=i3Xt5l9ybB2_nF_ma5cfPuL3OeFTs2dY2xDI0p4Azpg,
|
|
71
71
|
upgini/utils/track_info.py,sha256=G5Lu1xxakg2_TQjKZk4b5SvrHsATTXNVV3NbvWtT8k8,5663
|
72
72
|
upgini/utils/ts_utils.py,sha256=26vhC0pN7vLXK6R09EEkMK3Lwb9IVPH7LRdqFIQ3kPs,1383
|
73
73
|
upgini/utils/warning_counter.py,sha256=-GRY8EUggEBKODPSuXAkHn9KnEQwAORC0mmz_tim-PM,254
|
74
|
-
upgini-1.2.
|
75
|
-
upgini-1.2.
|
76
|
-
upgini-1.2.
|
77
|
-
upgini-1.2.
|
74
|
+
upgini-1.2.101.dist-info/METADATA,sha256=gHUIz6Ff8Poy8uSNBSEo3oXv674AkGAaissQL_HSp1g,49529
|
75
|
+
upgini-1.2.101.dist-info/WHEEL,sha256=zEMcRr9Kr03x1ozGwg5v9NQBKn3kndp6LSoSlVg-jhU,87
|
76
|
+
upgini-1.2.101.dist-info/licenses/LICENSE,sha256=5RRzgvdJUu3BUDfv4bzVU6FqKgwHlIay63pPCSmSgzw,1514
|
77
|
+
upgini-1.2.101.dist-info/RECORD,,
|
File without changes
|