upgini 1.2.130__py3-none-any.whl → 1.2.131__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 +1 -1
- upgini/utils/deduplicate_utils.py +5 -2
- {upgini-1.2.130.dist-info → upgini-1.2.131.dist-info}/METADATA +1 -1
- {upgini-1.2.130.dist-info → upgini-1.2.131.dist-info}/RECORD +7 -7
- {upgini-1.2.130.dist-info → upgini-1.2.131.dist-info}/WHEEL +0 -0
- {upgini-1.2.130.dist-info → upgini-1.2.131.dist-info}/licenses/LICENSE +0 -0
upgini/__about__.py
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
__version__ = "1.2.
|
|
1
|
+
__version__ = "1.2.131"
|
upgini/features_enricher.py
CHANGED
|
@@ -2786,7 +2786,7 @@ if response.status_code == 200:
|
|
|
2786
2786
|
df_without_features = df.drop(columns=features_not_to_pass, errors="ignore")
|
|
2787
2787
|
|
|
2788
2788
|
df_without_features, full_duplicates_warning = clean_full_duplicates(
|
|
2789
|
-
df_without_features, self.logger, bundle=self.bundle
|
|
2789
|
+
df_without_features, is_transform=True, logger=self.logger, bundle=self.bundle
|
|
2790
2790
|
)
|
|
2791
2791
|
if not silent_mode and full_duplicates_warning:
|
|
2792
2792
|
self.__log_warning(full_duplicates_warning)
|
|
@@ -160,7 +160,10 @@ def remove_fintech_duplicates(
|
|
|
160
160
|
|
|
161
161
|
|
|
162
162
|
def clean_full_duplicates(
|
|
163
|
-
df: pd.DataFrame,
|
|
163
|
+
df: pd.DataFrame,
|
|
164
|
+
is_transform: bool = False,
|
|
165
|
+
logger: Optional[Logger] = None,
|
|
166
|
+
bundle: Optional[ResourceBundle] = None,
|
|
164
167
|
) -> Tuple[pd.DataFrame, Optional[str]]:
|
|
165
168
|
if logger is None:
|
|
166
169
|
logger = logging.getLogger()
|
|
@@ -193,7 +196,7 @@ def clean_full_duplicates(
|
|
|
193
196
|
logger.warning(bundle.get("dataset_full_duplicates").format(share_full_dedup))
|
|
194
197
|
|
|
195
198
|
msg = None
|
|
196
|
-
if TARGET in df.columns:
|
|
199
|
+
if not is_transform and TARGET in df.columns:
|
|
197
200
|
unique_columns.remove(TARGET)
|
|
198
201
|
|
|
199
202
|
# Separate rows to exclude from deduplication:
|
|
@@ -1,9 +1,9 @@
|
|
|
1
|
-
upgini/__about__.py,sha256
|
|
1
|
+
upgini/__about__.py,sha256=-xQ6-PylS50NjUgd2Mk1UC4AWxxnecclkmuHdilo22Q,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=Nm2ZmwyQqvTnymYpGUwyJWy7y2ebXlHMyYmGeGcyA_s,31652
|
|
5
5
|
upgini/errors.py,sha256=2b_Wbo0OYhLUbrZqdLIx5jBnAsiD1Mcenh-VjR4HCTw,950
|
|
6
|
-
upgini/features_enricher.py,sha256=
|
|
6
|
+
upgini/features_enricher.py,sha256=Tw0LqlOPUWU0lMc3ekfW0qKV9iCyNtHS0dnQFAUNnxg,234396
|
|
7
7
|
upgini/http.py,sha256=-J_wOpnwVnT0ebPC6sOs6fN3AWtCD0LJLu6nlYmxaqk,44348
|
|
8
8
|
upgini/metadata.py,sha256=H3wiN37k-yqWZgbPD0tJzx8DzaCIkgmX5cybhByQWLg,12619
|
|
9
9
|
upgini/metrics.py,sha256=KCPE_apPN-9BIdv6GqASbJVaB_gBcy8wzNApAcyaGo4,46020
|
|
@@ -53,7 +53,7 @@ upgini/utils/country_utils.py,sha256=lY-eXWwFVegdVENFttbvLcgGDjFO17Sex8hd2PyJaRk
|
|
|
53
53
|
upgini/utils/custom_loss_utils.py,sha256=kieNZYBYZm5ZGBltF1F_jOSF4ea6C29rYuCyiDcqVNY,3857
|
|
54
54
|
upgini/utils/cv_utils.py,sha256=w6FQb9nO8BWDx88EF83NpjPLarK4eR4ia0Wg0kLBJC4,3525
|
|
55
55
|
upgini/utils/datetime_utils.py,sha256=l85UzSQLhtMeI2G6m-m8y8bCColCLSXNHb2-G6fKpLM,16988
|
|
56
|
-
upgini/utils/deduplicate_utils.py,sha256=
|
|
56
|
+
upgini/utils/deduplicate_utils.py,sha256=CLX0QapRxB-ZVQT7yGvv1vSd2zac5SwRjCJavujdCps,11332
|
|
57
57
|
upgini/utils/display_utils.py,sha256=p6o0VlYtGpU6bXv3B-fjQM9PeZEkl05OylHXSRyP0us,13219
|
|
58
58
|
upgini/utils/email_utils.py,sha256=pZ2vCfNxLIPUhxr0-OlABNXm12jjU44isBk8kGmqQzA,5277
|
|
59
59
|
upgini/utils/fallback_progress_bar.py,sha256=PDaKb8dYpVZaWMroNcOHsTc3pSjgi9mOm0--cOFTwJ0,1074
|
|
@@ -74,7 +74,7 @@ upgini/utils/target_utils.py,sha256=CihpV6SC95HwtlMH60rGAUzVDa4Id0Bva8ySprmNHlE,
|
|
|
74
74
|
upgini/utils/track_info.py,sha256=NDKeQTUlZaYp15UoP-xLKGoDoJQ0drbDMwB0g9R0PUg,6427
|
|
75
75
|
upgini/utils/ts_utils.py,sha256=26vhC0pN7vLXK6R09EEkMK3Lwb9IVPH7LRdqFIQ3kPs,1383
|
|
76
76
|
upgini/utils/warning_counter.py,sha256=-GRY8EUggEBKODPSuXAkHn9KnEQwAORC0mmz_tim-PM,254
|
|
77
|
-
upgini-1.2.
|
|
78
|
-
upgini-1.2.
|
|
79
|
-
upgini-1.2.
|
|
80
|
-
upgini-1.2.
|
|
77
|
+
upgini-1.2.131.dist-info/METADATA,sha256=qYItgAJjenacNSv7cAtVxbP1m53zLmFbv6qT-ReasgE,51142
|
|
78
|
+
upgini-1.2.131.dist-info/WHEEL,sha256=1yFddiXMmvYK7QYTqtRNtX66WJ0Mz8PYEiEUoOUUxRY,87
|
|
79
|
+
upgini-1.2.131.dist-info/licenses/LICENSE,sha256=5RRzgvdJUu3BUDfv4bzVU6FqKgwHlIay63pPCSmSgzw,1514
|
|
80
|
+
upgini-1.2.131.dist-info/RECORD,,
|
|
File without changes
|
|
File without changes
|