upgini 1.1.131a4__tar.gz → 1.1.132a1__tar.gz

{upgini-1.1.131a4/src/upgini.egg-info → upgini-1.1.132a1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: upgini
-Version: 1.1.131a4
+Version: 1.1.132a1
 Summary: Intelligent data search & enrichment for Machine Learning
 Home-page: https://upgini.com/
 Author: Upgini Developers

{upgini-1.1.131a4 → upgini-1.1.132a1}/setup.py

@@ -35,7 +35,7 @@ def send_log(msg: str):
 
 
 here = Path(__file__).parent.resolve()
-version = "1.1.131a4"
+version = "1.1.132a1"
 try:
     send_log(f"Start setup PyLib version {version}")
     setup(
@@ -77,7 +77,7 @@ try:
             "numpy>=1.19.0",
             "scikit-learn>=1.0.1",
             "pydantic>=1.8.2",
-            "fastparquet>=0.7.1",
+            "fastparquet>=0.8.1",
             "python-json-logger>=2.0.2",
             "catboost>=1.0.3",
             "lightgbm>=3.3.2",

{upgini-1.1.131a4 → upgini-1.1.132a1}/src/upgini/features_enricher.py

@@ -139,15 +139,16 @@ class FeaturesEnricher(TransformerMixin):
         raise_validation_error: bool = False,
         **kwargs,
     ):
-        self.api_key = api_key or os.environ.get(UPGINI_API_KEY)
+        self._api_key = api_key or os.environ.get(UPGINI_API_KEY)
         try:
-            self.rest_client = get_rest_client(endpoint, self.api_key)
+            self.rest_client = get_rest_client(endpoint, self._api_key)
         except UpginiConnectionError as e:
             print(e)
             return
 
+        self.logs_enabled = logs_enabled
         if logs_enabled:
-            self.logger = LoggerFactory().get_logger(endpoint, self.api_key)
+            self.logger = LoggerFactory().get_logger(endpoint, self._api_key)
         else:
             self.logger = logging.getLogger()
             self.logger.setLevel("FATAL")
@@ -174,7 +175,7 @@ class FeaturesEnricher(TransformerMixin):
             search_task = SearchTask(
                 search_id,
                 endpoint=self.endpoint,
-                api_key=self.api_key,
+                api_key=self._api_key,
             )
 
             print(bundle.get("search_by_task_id_start"))
@@ -235,6 +236,16 @@ class FeaturesEnricher(TransformerMixin):
         self.__cached_sampled_datasets: Optional[Tuple[pd.DataFrame, pd.DataFrame, pd.Series, Dict, Dict]] = None
         self.raise_validation_error = raise_validation_error
 
+    def _get_api_key(self):
+        return self._api_key
+
+    def _set_api_key(self, api_key: str):
+        self._api_key = api_key
+        if self.logs_enabled:
+            self.logger = LoggerFactory().get_logger(self.endpoint, self._api_key)
+
+    api_key = property(_get_api_key, _set_api_key)
+
     def fit(
         self,
         X: Union[pd.DataFrame, pd.Series, np.ndarray],
@@ -579,13 +590,14 @@ class FeaturesEnricher(TransformerMixin):
             finally:
                 self.logger.info(f"Transform elapsed time: {time.time() - start_time}")
 
-            if self.country_added:
-                result = drop_existing_columns(result, COUNTRY)
+            if result is not None:
+                if self.country_added:
+                    result = drop_existing_columns(result, COUNTRY)
 
-            if keep_input:
-                return result
-            else:
-                return drop_existing_columns(result, X.columns)
+                if keep_input:
+                    return result
+                else:
+                    return drop_existing_columns(result, X.columns)
 
     def calculate_metrics(
         self,
@@ -912,13 +924,21 @@ class FeaturesEnricher(TransformerMixin):
     def _has_features_with_commercial_schema(
         self, commercial_schema: str, exclude_features_sources: Optional[List[str]]
     ) -> bool:
+        return len(self._get_features_with_commercial_schema(commercial_schema, exclude_features_sources)) > 0
+
+    def _get_features_with_commercial_schema(
+        self, commercial_schema: str, exclude_features_sources: Optional[List[str]]
+    ) -> List[str]:
         if exclude_features_sources:
             filtered_features_info = self.features_info[
                 ~self.features_info[bundle.get("features_info_name")].isin(exclude_features_sources)
             ]
         else:
             filtered_features_info = self.features_info
-        return (filtered_features_info[bundle.get("features_info_commercial_schema")] == commercial_schema).any()
+        return list(filtered_features_info.loc[
+            filtered_features_info[bundle.get("features_info_commercial_schema")] == commercial_schema,
+            bundle.get("features_info_name"),
+        ].values)
 
     def _has_trial_features(self, exclude_features_sources: Optional[List[str]]) -> bool:
         return self._has_features_with_commercial_schema(CommercialSchema.TRIAL.value, exclude_features_sources)
@@ -1248,7 +1268,6 @@ class FeaturesEnricher(TransformerMixin):
                 msg = bundle.get("transform_with_trial_features")
                 self.logger.warn(msg)
                 print(msg)
-                return None
 
             columns_to_drop = [c for c in validated_X.columns if c in self.feature_names_]
             if len(columns_to_drop) > 0:
@@ -1493,7 +1512,8 @@ class FeaturesEnricher(TransformerMixin):
         if is_demo_dataset:
             msg = bundle.get("demo_dataset_info")
             self.logger.info(msg)
-            print(msg)
+            if not self.__is_registered:
+                print(msg)
 
         if self.generate_features is not None and len(self.generate_features) > 0:
             x_columns = list(validated_X.columns)

{upgini-1.1.131a4 → upgini-1.1.132a1}/src/upgini/http.py

@@ -11,8 +11,8 @@ from http.client import HTTPConnection
 from json import dumps
 from typing import Dict, List, Optional
 from urllib.parse import urljoin
-import pandas as pd
 
+import pandas as pd
 import requests
 from pydantic import BaseModel
 from pythonjsonlogger import jsonlogger
@@ -355,7 +355,9 @@ class _RestClient:
                 digest = md5_hash.hexdigest()
                 metadata_with_md5 = metadata.copy(update={"checksumMD5": digest})
 
-            digest_sha256 = hashlib.sha256(pd.util.hash_pandas_object(pd.read_parquet(file_path)).values).hexdigest()
+            digest_sha256 = hashlib.sha256(
+                pd.util.hash_pandas_object(pd.read_parquet(file_path)).values
+            ).hexdigest()
             metadata_with_md5 = metadata_with_md5.copy(update={"digest": digest_sha256})
 
             with open(file_path, "rb") as file:
@@ -433,8 +435,9 @@ class _RestClient:
                 digest = md5_hash.hexdigest()
                 metadata_with_md5 = metadata.copy(update={"checksumMD5": digest})
 
-            import pandas as pd
-            digest_sha256 = hashlib.sha256(pd.util.hash_pandas_object(pd.read_parquet(file_path)).values).hexdigest()
+            digest_sha256 = hashlib.sha256(
+                pd.util.hash_pandas_object(pd.read_parquet(file_path)).values
+            ).hexdigest()
             metadata_with_md5 = metadata_with_md5.copy(update={"digest": digest_sha256})
 
             with open(file_path, "rb") as file:
@@ -787,17 +790,18 @@ class BackendLogHandler(logging.Handler):
     def __init__(self, rest_client: _RestClient, *args, **kwargs) -> None:
         super().__init__(*args, **kwargs)
         self.rest_client = rest_client
-        print("Before track metrics")
-        self.track_metrics = get_track_metrics()
-        print("After track metrics")
-        if "ip" in self.track_metrics.keys():
-            self.hostname = self.track_metrics["ip"]
-        else:
-            self.hostname = "0.0.0.0"
+        self.track_metrics = None
+        self.hostname = None
 
     def emit(self, record: logging.LogRecord) -> None:
         def task():
             try:
+                if self.track_metrics is None:
+                    self.track_metrics = get_track_metrics()
+                    if "ip" in self.track_metrics.keys():
+                        self.hostname = self.track_metrics["ip"]
+                    else:
+                        self.hostname = "0.0.0.0"
                 text = self.format(record)
                 tags = self.track_metrics
                 tags["version"] = __version__
@@ -848,11 +852,8 @@ class LoggerFactory:
 
         upgini_logger = logging.getLogger(f"upgini.{hash(key)}")
         upgini_logger.handlers.clear()
-        print("Before rest client")
         rest_client = get_rest_client(backend_url, api_token)
-        print("Before backend log handler")
         datadog_handler = BackendLogHandler(rest_client)
-        print("After backend log handler")
         json_formatter = jsonlogger.JsonFormatter(
             "%(asctime)s %(threadName)s %(name)s %(levelname)s %(message)s",
             timestamp=True,

{upgini-1.1.131a4 → upgini-1.1.132a1}/src/upgini/resource_bundle/strings.properties

@@ -25,7 +25,8 @@ metrics_no_important_free_features=WARNING: No important free features to calcul
 metrics_no_important_features=WARNING: No important features to calculate metrics
 metrics_negative_uplift_without_cv=Please re-check that your task is not a time series prediction. If so, restart search with cv=CVType.time_series param for correct search results. See docs https://github.com/upgini/upgini#-time-series-prediction-support
 metrics_with_trial_features=The calculation of final accuracy metrics using Trial data is not available for unauthorized users.\nGet a free API key on https://upgini.com and repeat your request.
-transform_with_trial_features=Enriching with Trial data is not available for unauthorized users.\nGet a free API key on https://upgini.com and repeat your request.
+transform_with_trial_features=WARNING: Your search results contain Trial data sources. To enrich your dataframe using transform or fit_transform with features from these Trial data sources, please register for a Free API key at https://upgini.com and resubmit your request.
+# Enriching with Trial data is not available for unauthorized users.\nGet a free API key on https://upgini.com and repeat your request.
 metrics_with_paid_features=The calculation of final accuracy metrics using Paid data is not available.\nContact Upgini support for the data access
 transform_with_paid_features=Enriching with Paid data is not available.\nContact Upgini support for the data access
 trial_quota_limit_riched=WARNING: You have reached the quota limit of trial data usage. Please contact Upgini support to remove restriction

{upgini-1.1.131a4 → upgini-1.1.132a1}/src/upgini/search_task.py

@@ -59,8 +59,8 @@ class SearchTask:
         submitted_statuses = {"SUBMITTED", "VALIDATION_SUBMITTED"}
         if not quiet:
             print(bundle.get("polling_search_task").format(self.search_task_id))
-        if is_demo_api_key(self.api_key):
-            print(bundle.get("polling_unregister_information"))
+            if is_demo_api_key(self.api_key):
+                print(bundle.get("polling_unregister_information"))
         search_task_id = self.initial_search_task_id if self.initial_search_task_id is not None else self.search_task_id
 
         try:

{upgini-1.1.131a4 → upgini-1.1.132a1/src/upgini.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: upgini
-Version: 1.1.131a4
+Version: 1.1.132a1
 Summary: Intelligent data search & enrichment for Machine Learning
 Home-page: https://upgini.com/
 Author: Upgini Developers

{upgini-1.1.131a4 → upgini-1.1.132a1}/src/upgini.egg-info/requires.txt

@@ -4,7 +4,7 @@ pandas<2.0.0,>=1.1.0
 numpy>=1.19.0
 scikit-learn>=1.0.1
 pydantic>=1.8.2
-fastparquet>=0.7.1
+fastparquet>=0.8.1
 python-json-logger>=2.0.2
 catboost>=1.0.3
 lightgbm>=3.3.2

{upgini-1.1.131a4 → upgini-1.1.132a1}/tests/test_features_enricher.py

@@ -397,7 +397,7 @@ def test_saved_features_enricher(requests_mock: Mocker):
             {
                 "segment": [train_segment, eval_1_segment, eval_2_segment],
                 rows_header: [10000, 1000, 1000],
-                enriched_rocauc: [0.507052, 0.485257, 0.491804],
+                enriched_rocauc: [0.500276, 0.499805, 0.497979],
             }
         )
         .set_index("segment")
@@ -1774,7 +1774,7 @@ def test_correct_order_of_enriched_X(requests_mock: Mocker):
     df_with_eval_set_index_with_date = converter.convert(df_with_eval_set_index)
     mock_features["system_record_id"] = pd.util.hash_pandas_object(
         df_with_eval_set_index_with_date[sorted(search_keys.keys())].reset_index(drop=True), index=False
-    )
+    ).astype("Float64")
     mock_validation_raw_features(requests_mock, url, validation_search_task_id, mock_features)
 
     enriched_df_with_eval_set = enricher.transform(df_with_eval_set_index)
@@ -2415,12 +2415,12 @@ def test_diff_target_dups(requests_mock: Mocker):
         self.validate()
         assert len(self.data) == 2
         print(self.data)
-        assert self.data.loc[2, "date_fake_a"] == 1672531200000
-        assert self.data.loc[2, "feature_fake_a"] == 12
-        assert self.data.loc[2, "target"] == 0
-        assert self.data.loc[3, "date_fake_a"] == 1672531200000
-        assert self.data.loc[3, "feature_fake_a"] == 13
-        assert self.data.loc[3, "target"] == 1
+        assert self.data.loc[0, "date_fake_a"] == 1672531200000
+        assert self.data.loc[0, "feature_fake_a"] == 12
+        assert self.data.loc[0, "target"] == 0
+        assert self.data.loc[1, "date_fake_a"] == 1672531200000
+        assert self.data.loc[1, "feature_fake_a"] == 13
+        assert self.data.loc[1, "target"] == 1
         return SearchTask("123", self, endpoint=url, api_key="fake_api_key")
 
     Dataset.search = mock_search