upgini 1.1.103__py3-none-any.whl → 1.1.104__py3-none-any.whl

upgini/data_source/data_source_publisher.py

@@ -44,6 +44,7 @@ class DataSourcePublisher:
         exclude_columns: Optional[List[str]] = None,
         hash_feature_names=False,
         snapshot_frequency_days: Optional[int] = None,
+        features_for_embeddings: Optional[List[str]] = None,
     ) -> str:
         trace_id = str(uuid.uuid4())
 
@@ -65,6 +66,7 @@ class DataSourcePublisher:
                     "excludeColumns": exclude_columns,
                     "hashFeatureNames": hash_feature_names,
                     "snapshotFrequencyDays": snapshot_frequency_days,
+                    "featuresForEmbeddings": features_for_embeddings,
                 }
                 self.logger.info(f"Start registering data table {request}")
 

upgini/features_enricher.py

@@ -639,8 +639,8 @@ class FeaturesEnricher(TransformerMixin):
                 if (
                     self._search_task is None
                     or self._search_task.initial_max_hit_rate_v2() is None
-                    or self.X is None
-                    or self.y is None
+                    or (self.X is None and X is None)
+                    or (self.y is None and y is None)
                 ):
                     raise ValidationError(bundle.get("metrics_unfitted_enricher"))
 
@@ -653,6 +653,18 @@ class FeaturesEnricher(TransformerMixin):
                     self.__display_slack_community_link(msg)
                     return None
 
+                if (
+                    estimator is not None
+                    and hasattr(estimator, "get_param")
+                    and estimator.get_param("cat_features") is not None
+                ):
+                    cat_features = estimator.get_param("cat_features")
+                    if len(cat_features) > 0 and isinstance(cat_features[0], int):
+                        effectiveX = X or self.X
+                        cat_features = [effectiveX.columns[i] for i in cat_features]
+                else:
+                    cat_features = None
+
                 prepared_data = self._prepare_data_for_metrics(
                     trace_id, X, y, eval_set, exclude_features_sources, importance_threshold, max_features
                 )
@@ -699,7 +711,7 @@ class FeaturesEnricher(TransformerMixin):
                             f"Calculate baseline {metric} on client features: {fitting_X.columns.to_list()}"
                         )
                         baseline_estimator = EstimatorWrapper.create(
-                            estimator, self.logger, model_task_type, _cv, fitting_enriched_X, scoring
+                            estimator, self.logger, model_task_type, _cv, fitting_enriched_X, scoring, cat_features
                         )
                         etalon_metric = baseline_estimator.cross_val_predict(fitting_X, y_sorted)
 
@@ -711,7 +723,7 @@ class FeaturesEnricher(TransformerMixin):
                             f"Calculate enriched {metric} on combined features: {fitting_enriched_X.columns.to_list()}"
                         )
                         enriched_estimator = EstimatorWrapper.create(
-                            estimator, self.logger, model_task_type, _cv, fitting_enriched_X, scoring
+                            estimator, self.logger, model_task_type, _cv, fitting_enriched_X, scoring, cat_features
                         )
                         enriched_metric = enriched_estimator.cross_val_predict(fitting_enriched_X, enriched_y_sorted)
                         if etalon_metric is not None:

upgini/metrics.py

@@ -1,5 +1,5 @@
 import logging
-from typing import Callable, List, Tuple, Union
+from typing import Callable, List, Optional, Tuple, Union
 
 import numpy as np
 import pandas as pd
@@ -14,6 +14,7 @@ from sklearn.metrics._regression import (
     mean_squared_error,
 )
 from sklearn.model_selection import BaseCrossValidator, cross_validate
+from copy import deepcopy
 
 from upgini.errors import ValidationError
 from upgini.metadata import ModelTaskType
@@ -40,7 +41,7 @@ LIGHTGBM_PARAMS = {
     "max_depth": 4,
     "n_estimators": 150,
     "learning_rate": 0.05,
-    "min_child_weight": 1
+    "min_child_weight": 1,
 }
 
 N_FOLDS = 5
@@ -129,6 +130,7 @@ class EstimatorWrapper:
         cv: BaseCrossValidator,
         X: pd.DataFrame,
         scoring: Union[Callable, str, None] = None,
+        cat_features: Optional[List[str]] = None,
     ) -> "EstimatorWrapper":
         scorer, metric_name, multiplier = _get_scorer(target_type, scoring)
         kwargs = {
@@ -149,8 +151,16 @@ class EstimatorWrapper:
             else:
                 raise Exception(bundle.get("metrics_unsupported_target_type").format(target_type))
         else:
-            kwargs["estimator"] = estimator
+            if hasattr(estimator, "copy"):
+                estimator_copy = estimator.copy()
+            else:
+                estimator_copy = deepcopy(estimator)
+            kwargs["estimator"] = estimator_copy
             if isinstance(estimator, CatBoostClassifier) or isinstance(estimator, CatBoostRegressor):
+                if cat_features is not None:
+                    estimator_copy.set_params(cat_features=[
+                        X.columns.get_loc(cat_feature) for cat_feature in cat_features
+                    ])
                 estimator = CatBoostWrapper(**kwargs)
             else:
                 try:
@@ -197,6 +207,15 @@ class CatBoostWrapper(EstimatorWrapper):
             else:
                 X = X.drop(columns=name)
         cat_features_idx = [X.columns.get_loc(c) for c in unique_cat_features]
+        if (
+            hasattr(self.estimator, "get_param")
+            and hasattr(self.estimator, "_init_params")
+            and self.estimator.get_param("cat_features") is not None
+        ):
+            cat_features_set = set(cat_features_idx)
+            cat_features_set.update(self.estimator.get_param("cat_features"))
+            cat_features_idx = list(cat_features_set)
+            del self.estimator._init_params["cat_features"]
 
         params.update({"cat_features": cat_features_idx})
         return X, y, params
@@ -376,6 +395,6 @@ def _ext_mean_squared_log_error(y_true, y_pred, *, sample_weight=None, multioutp
 def _is_too_many_categorical_values(X: pd.DataFrame) -> bool:
     many_values_features_count = 0
     for f in _get_cat_features(X):
-        if X[f].nunique() > 100:
+        if X[f].astype("string").nunique() > 100:
             many_values_features_count += 1
     return many_values_features_count >= 2

upgini/utils/email_utils.py

@@ -9,8 +9,7 @@ from pandas.api.types import is_string_dtype
 from upgini.metadata import SearchKey
 from upgini.utils.base_search_key_detector import BaseSearchKeyDetector
 
-
-EMAIL_REGEX = re.compile(r'([A-Za-z0-9]+[.-_])*[A-Za-z0-9]+@[A-Za-z0-9-]+(\.[A-Z|a-z]{2,})+')
+EMAIL_REGEX = re.compile(r"^[a-zA-Z0-9.!#$%&’*+/=?^_`{|}~-]+@[a-zA-Z0-9-]+(?:\.[a-zA-Z0-9-]+)*$")
 
 
 class EmailSearchKeyDetector(BaseSearchKeyDetector):
@@ -20,13 +19,11 @@ class EmailSearchKeyDetector(BaseSearchKeyDetector):
     def _is_search_key_by_values(self, column: pd.Series) -> bool:
         if not is_string_dtype(column):
             return False
+        if not column.astype("string").str.contains("@").any():
+            return False
 
         all_count = len(column)
-        is_email_count = len(
-            column.loc[
-                column.astype("string").str.fullmatch(EMAIL_REGEX)
-            ]
-        )
+        is_email_count = len(column.loc[column.astype("string").str.fullmatch(EMAIL_REGEX)])
         return is_email_count / all_count > 0.1
 
 

{upgini-1.1.103.dist-info → upgini-1.1.104.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: upgini
-Version: 1.1.103
+Version: 1.1.104
 Summary: Low-code feature search and enrichment library for machine learning
 Home-page: https://upgini.com/
 Author: Upgini Developers

{upgini-1.1.103.dist-info → upgini-1.1.104.dist-info}/RECORD

@@ -2,17 +2,17 @@ upgini/__init__.py,sha256=asENHgEVHQBIkV-e_0IhE_ZWqkCG6398U3ZLrNzAH6k,407
 upgini/ads.py,sha256=mre6xn44wcC_fg63iLT_kTh4mViZqR9AKRJZAtpQz8Y,2592
 upgini/dataset.py,sha256=fl01WTbM2smgskjrHQJS9oTzymEj5ZulGngCU_d5PnQ,42110
 upgini/errors.py,sha256=BqpvfhW2jJW5fa5KXj0alhXatGl-WK4xTl309-QNLp8,959
-upgini/features_enricher.py,sha256=bjPiQDGn2ULS1dEFq7vIFxF49QM8UjJWFvTewqjDdHw,109147
+upgini/features_enricher.py,sha256=PtZIiOtpLItYRm3U7e5gsWAwAiTze4rznuKFFHjFpuQ,109768
 upgini/http.py,sha256=kgWj6wU1PbGPoGAbRvK35umXQ5zwEfEKeGy5Az0fss0,35479
 upgini/metadata.py,sha256=GPGsaGi5UtePQR2Qiqc7OJZn-ewvHmvepn3P_wJDW7Y,5856
-upgini/metrics.py,sha256=3gKDUJe4IzcS32hLitbPj2G-y-F66eyzKw0DWSIkun0,13937
+upgini/metrics.py,sha256=uJhtGKgUUFnvdF16xscfe9AGDoDN6LqUV97RWDP39NU,14869
 upgini/search_task.py,sha256=H7l-BhCRF9t58D0L1xNdC_qU_JFHYnAZZ165fVDQgmM,33884
 upgini/spinner.py,sha256=X9a0xhj0QVIwjVTTjXUTuAgPBnyrLbW-B6G534fxs1E,1149
 upgini/version_validator.py,sha256=rDIncP6BEko4J2F2hUcMOtKm_vZbI4ICWcNcw8hrwM4,1400
 upgini/ads_management/__init__.py,sha256=qzyisOToVRP-tquAJD1PblZhNtMrOB8FiyF9JvfkvgE,50
 upgini/ads_management/ads_manager.py,sha256=Cc3v4lLLpM0g4oUH_q2DYFN3bNWpSmltAGnZQby3G74,2630
 upgini/data_source/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-upgini/data_source/data_source_publisher.py,sha256=siGi3jWuP1wu4QS5g5XwNYGhCF0ILOw14qRXCfxD2jo,8415
+upgini/data_source/data_source_publisher.py,sha256=-Tpqiw6xrCinxdDKIEg6aS68ZqLwxoBrg4J4PTQNs6g,8546
 upgini/mdc/__init__.py,sha256=CuKmWYCqAnmiq1S7wgMzJhSCTsXuoeiZWXSfzw0lyig,1152
 upgini/mdc/context.py,sha256=eVNEubcgkiAP139Vna2qtUBZJWoy15rWWAuB0TFv54E,1484
 upgini/normalizer/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -31,7 +31,7 @@ upgini/utils/country_utils.py,sha256=9BXSXoGm3nVoOZE_bRENY-KMkwMUFvAF3Au0zxUNA1o
 upgini/utils/cv_utils.py,sha256=PeexQVPWrpUNlmwGtfU1FWA-aI1UyrMDgMT594ErpxA,2252
 upgini/utils/datetime_utils.py,sha256=PK1Fc5rJ_UhCJc1TNOZPSrtsYxjD7v9dsBYOZj1RKvo,4292
 upgini/utils/display_utils.py,sha256=iG3-hdv8_rJDWKwnQYIi1SHF-gLPAEi8jjk_05-qtMg,1934
-upgini/utils/email_utils.py,sha256=H05wMKVZML36Ipwxv1C4StDUyPJVaXWgU5NA06UDJMo,3048
+upgini/utils/email_utils.py,sha256=2IUxP1e8DsmU4qS1BN3n1JmuziZO_cV35fNf4Di0yxc,3090
 upgini/utils/features_validator.py,sha256=LZAKTWtmINWII09UHF0R0muEz7yHLGlJkLUk8zM305Q,2190
 upgini/utils/format.py,sha256=Yv5cvvSs2bOLUzzNu96Pu33VMDNbabio92QepUj41jU,243
 upgini/utils/phone_utils.py,sha256=JNSkF8G6mgsN8Czy11pamaJdsY6rBINEMpi7jbVt_RA,408
@@ -39,8 +39,8 @@ upgini/utils/postal_code_utils.py,sha256=_8CR9tBqsPptQsmMUvnrCAmBaMIQSWH3JfJ4ly3
 upgini/utils/target_utils.py,sha256=3eHrDy_Dc9ozuOwHGnGA705m9glCxKmjB-DfLrflqiA,1370
 upgini/utils/track_info.py,sha256=O_oL4gy1jH0DVgtiUeZAW0YKCeRT4B_bzH_SZYkFaOE,4076
 upgini/utils/warning_counter.py,sha256=vnmdFo5-7GBkU2bK9h_uC0K0Y_wtfcYstxOdeRfacO0,228
-upgini-1.1.103.dist-info/LICENSE,sha256=5RRzgvdJUu3BUDfv4bzVU6FqKgwHlIay63pPCSmSgzw,1514
-upgini-1.1.103.dist-info/METADATA,sha256=8zAaEAC2-WpBOPqrRagZYdO1mABJzDTWhVgs7R86E1Q,41101
-upgini-1.1.103.dist-info/WHEEL,sha256=G16H4A3IeoQmnOrYV4ueZGKSjhipXx8zc8nu9FGlvMA,92
-upgini-1.1.103.dist-info/top_level.txt,sha256=OFhTGiDIWKl5gFI49qvWq1R9IKflPaE2PekcbDXDtx4,7
-upgini-1.1.103.dist-info/RECORD,,
+upgini-1.1.104.dist-info/LICENSE,sha256=5RRzgvdJUu3BUDfv4bzVU6FqKgwHlIay63pPCSmSgzw,1514
+upgini-1.1.104.dist-info/METADATA,sha256=nvFzylBFXaBafhSec5_Ja5KfVcrNc2pzAHNRVzBMzhA,41101
+upgini-1.1.104.dist-info/WHEEL,sha256=G16H4A3IeoQmnOrYV4ueZGKSjhipXx8zc8nu9FGlvMA,92
+upgini-1.1.104.dist-info/top_level.txt,sha256=OFhTGiDIWKl5gFI49qvWq1R9IKflPaE2PekcbDXDtx4,7
+upgini-1.1.104.dist-info/RECORD,,