upgini 1.2.99a3922.dev5__tar.gz → 1.2.103__tar.gz

{upgini-1.2.99a3922.dev5 → upgini-1.2.103}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: upgini
-Version: 1.2.99a3922.dev5
+Version: 1.2.103
 Summary: Intelligent data search & enrichment for Machine Learning
 Project-URL: Bug Reports, https://github.com/upgini/upgini/issues
 Project-URL: Homepage, https://upgini.com/

upgini-1.2.103/src/upgini/__about__.py

@@ -0,0 +1 @@
+__version__ = "1.2.103"

{upgini-1.2.99a3922.dev5 → upgini-1.2.103}/src/upgini/autofe/feature.py

@@ -164,6 +164,7 @@ class Feature:
 
     def delete_data(self):
         self.data = None
+        self.op.delete_data()
         for child in self.children:
             child.delete_data()
 

{upgini-1.2.99a3922.dev5 → upgini-1.2.103}/src/upgini/autofe/operator.py

@@ -86,6 +86,9 @@ class Operator(BaseModel, metaclass=OperatorRegistry):
         self._logger = logger
         return self
 
+    def delete_data(self):
+        pass
+
 
 class ParametrizedOperator(Operator, abc.ABC):
 

upgini-1.2.103/src/upgini/autofe/vector.py

@@ -0,0 +1,75 @@
+from typing import List, Optional
+
+import pandas as pd
+
+from upgini.autofe.operator import OperatorRegistry, PandasOperator, ParametrizedOperator, VectorizableMixin
+
+
+class Mean(PandasOperator, VectorizableMixin):
+    name: str = "mean"
+    output_type: Optional[str] = "float"
+    is_vector: bool = True
+    group_index: int = 0
+
+    def calculate_vector(self, data: List[pd.Series]) -> pd.Series:
+        return pd.DataFrame(data).T.fillna(0).mean(axis=1)
+
+
+class Sum(PandasOperator, VectorizableMixin):
+    name: str = "sum"
+    is_vector: bool = True
+    group_index: int = 0
+
+    def calculate_vector(self, data: List[pd.Series]) -> pd.Series:
+        return pd.DataFrame(data).T.fillna(0).sum(axis=1)
+
+
+class Vectorize(PandasOperator, VectorizableMixin):
+    name: str = "vectorize"
+    is_vector: bool = True
+    group_index: int = 0
+
+    def calculate_vector(self, data: List[pd.Series]) -> pd.Series:
+        return pd.DataFrame(data).T.apply(lambda x: x.to_list(), axis=1)
+
+
+class OnnxModel(PandasOperator, ParametrizedOperator, metaclass=OperatorRegistry):
+    name: str = "onnx"
+    score_name: str = "score"
+    is_vector: bool = True
+    output_type: Optional[str] = "float"
+
+    def to_formula(self) -> str:
+        return f"onnx_{self.score_name}"
+
+    @classmethod
+    def from_formula(cls, formula: str) -> Optional["OnnxModel"]:
+        if "(" in formula:
+            return None
+        if formula.startswith("onnx_"):
+            score_name = formula[len("onnx_"):]
+            return cls(score_name=score_name)
+        elif formula == "onnx":  # for OperatorRegistry
+            return cls()
+        return None
+
+
+class CatboostModel(PandasOperator, ParametrizedOperator, metaclass=OperatorRegistry):
+    name: str = "catboost"
+    score_name: str = "score"
+    is_vector: bool = True
+    output_type: Optional[str] = "float"
+
+    def to_formula(self) -> str:
+        return f"catboost_{self.score_name}"
+
+    @classmethod
+    def from_formula(cls, formula: str) -> Optional["CatboostModel"]:
+        if "(" in formula:
+            return None
+        if formula.startswith("catboost_"):
+            score_name = formula[len("catboost_"):]
+            return cls(score_name=score_name)
+        elif formula == "catboost":  # for OperatorRegistry
+            return cls()
+        return None

{upgini-1.2.99a3922.dev5 → upgini-1.2.103}/src/upgini/data_source/data_source_publisher.py

@@ -500,6 +500,8 @@ class DataSourcePublisher:
         name: str,
         input_names: List[str],
         search_id: str,
+        date_column: Optional[str] = None,
+        score_name: Optional[str] = None,
         model_type: Optional[Literal["ONNX", "CATBOOST"]] = None,
         description: str = "",
     ):
@@ -508,6 +510,8 @@ class DataSourcePublisher:
         metadata = {
             "modelName": name,
             "inputNames": input_names,
+            "dateColumn": date_column,
+            "scoreName": score_name,
             "searchTaskId": search_id,
             "modelType": model_type or "ONNX",
             "description": description,

{upgini-1.2.99a3922.dev5 → upgini-1.2.103}/src/upgini/features_enricher.py

@@ -3708,6 +3708,11 @@ if response.status_code == 200:
 
     @staticmethod
     def _get_group_columns(df: pd.DataFrame, search_keys: Dict[str, SearchKey]) -> List[str]:
+        search_key_priority = [SearchKey.PHONE, SearchKey.EMAIL, SearchKey.HEM, SearchKey.IP]
+        for key_type in search_key_priority:
+            if key_type in search_keys.values():
+                return [col for col, t in search_keys.items() if t == key_type]
+
         return [
             col
             for col, t in search_keys.items()
@@ -4481,7 +4486,9 @@ if response.status_code == 200:
         sample = df.head(100)
 
         def check_need_detect(search_key: SearchKey):
-            return not is_transform or search_key in self.fit_search_keys.values()
+            return not is_transform or (
+                search_key in self.fit_search_keys.values() and search_key not in search_keys.values()
+            )
 
         # if SearchKey.POSTAL_CODE not in search_keys.values() and check_need_detect(SearchKey.POSTAL_CODE):
         if check_need_detect(SearchKey.POSTAL_CODE):

{upgini-1.2.99a3922.dev5 → upgini-1.2.103}/src/upgini/resource_bundle/strings.properties

@@ -99,6 +99,7 @@ unsupported_date_type=Unsupported type of date column `{}`. Convert to datetime
 invalid_postal_code=All values of POSTAL_CODE column `{}` are invalid
 invalid_country=All values of COUNTRY column `{}` are invalid
 invalid_ip=All values of IP column `{}` are invalid
+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.
     # X and y validation
 unsupported_x_type=Unsupported type of X: {}. Use pandas.DataFrame, pandas.Series or numpy.ndarray or list
 x_contains_dup_columns=X contains duplicate column names. Please rename or drop duplicates

{upgini-1.2.99a3922.dev5 → upgini-1.2.103}/src/upgini/utils/display_utils.py

@@ -152,10 +152,6 @@ def make_html_report(
     email: Optional[str] = None,
     search_keys: Optional[List[str]] = None,
 ) -> str:
-    # relevant_features_df = relevant_features_df.copy()
-    # relevant_features_df["Feature name"] = relevant_features_df["Feature name"].apply(
-    #     lambda x: "*" + x if x.contains("_autofe_") else x
-    # )
     relevant_datasources_df = relevant_datasources_df.copy()
     relevant_datasources_df["action"] = (
         f"""<a href="https://upgini.com/request-a-quote?search-id={search_id}">"""

upgini-1.2.99a3922.dev5/src/upgini/__about__.py

@@ -1 +0,0 @@
-__version__ = "1.2.99a3922.dev5"

upgini-1.2.99a3922.dev5/src/upgini/autofe/vector.py

@@ -1,45 +0,0 @@
-from typing import List, Optional
-
-import pandas as pd
-
-from upgini.autofe.operator import OperatorRegistry, PandasOperator, VectorizableMixin
-
-
-class Mean(PandasOperator, VectorizableMixin):
-    name: str = "mean"
-    output_type: Optional[str] = "float"
-    is_vector: bool = True
-    group_index: int = 0
-
-    def calculate_vector(self, data: List[pd.Series]) -> pd.Series:
-        return pd.DataFrame(data).T.fillna(0).mean(axis=1)
-
-
-class Sum(PandasOperator, VectorizableMixin):
-    name: str = "sum"
-    is_vector: bool = True
-    group_index: int = 0
-
-    def calculate_vector(self, data: List[pd.Series]) -> pd.Series:
-        return pd.DataFrame(data).T.fillna(0).sum(axis=1)
-
-
-class Vectorize(PandasOperator, VectorizableMixin):
-    name: str = "vectorize"
-    is_vector: bool = True
-    group_index: int = 0
-
-    def calculate_vector(self, data: List[pd.Series]) -> pd.Series:
-        return pd.DataFrame(data).T.apply(lambda x: x.to_list(), axis=1)
-
-
-class OnnxModel(PandasOperator, metaclass=OperatorRegistry):
-    name: str = "onnx"
-    is_vector: bool = True
-    output_type: Optional[str] = "float"
-
-
-class CatboostModel(PandasOperator, metaclass=OperatorRegistry):
-    name: str = "catboost"
-    is_vector: bool = True
-    output_type: Optional[str] = "float"