Merge remote-tracking branch 'origin/master'

# Conflicts:
#	hypex/comparators/__init__.py

Author: tikhomirovd

examples/experiments/performance_test/performance_test.py

@@ -1 +1 @@
-__version__ = "1.0.1"
+__version__ = "1.0.2"

hypex/__version__.py

@@ -117,7 +117,7 @@ def execute(self, data: ExperimentData) -> ExperimentData:
                 for i in range(0, len(analysis_ids), len(analysis_ids) // num_groups):
                     value = t_data.iloc[i : i + len(analysis_ids) // num_groups][f]
                     multitest_pvalues = self._add_pvalues(multitest_pvalues, value, f)
-                    analysis_data[f"{c} {f} {groups[i//num_groups + 1][0]}"] = (
+                    analysis_data[f"{c} {f} {groups[i // num_groups + 1][0]}"] = (
                         value.mean()
                     )
             if c not in ["UTest", "TTest"]:
@@ -126,7 +126,7 @@ def execute(self, data: ExperimentData) -> ExperimentData:
                 for idx, value in zip(indexes, values):
                     name = idx.split(ID_SPLIT_SYMBOL)[-1]
                     analysis_data[
-                        f"{c} {name[name.find(NAME_BORDER_SYMBOL) + 1: name.rfind(NAME_BORDER_SYMBOL)]}"
+                        f"{c} {name[name.find(NAME_BORDER_SYMBOL) + 1 : name.rfind(NAME_BORDER_SYMBOL)]}"
                     ] = value[0]
 
         analysis_dataset = Dataset.from_dict(

hypex/analyzers/ab.py

@@ -134,23 +134,20 @@ def _extract_dataset(
             return result
         return Dataset.from_dict(compare_result, roles, BackendsEnum.pandas)
 
-    # TODO выделить в отдельную функцию с кваргами (нужно для альфы)
-
     @staticmethod
     def _grouping_data_split(
         grouping_data: dict[str, Dataset],
         compare_by: Literal["groups", "columns", "columns_in_groups", "cross"],
         target_fields: list[str],
         baseline_field: str | None = None,
     ) -> GroupingDataType:
-        if isinstance(grouping_data, dict):
-            compared_data = [(name, data) for name, data in grouping_data.items()]
-            baseline_data = [compared_data.pop(0)]
-        else:
+        if not isinstance(grouping_data, dict):
             raise TypeError(
                 f"Grouping data must be dict of strings and datasets, but got {type(grouping_data)}"
             )
 
+        compared_data = list(grouping_data.items())
+        baseline_data = [compared_data.pop(0)]
         baseline_data = [
             (
                 bucket[0],
@@ -166,7 +163,7 @@ def _grouping_data_split(
 
     @staticmethod
     def _split_ds_into_columns(
-        data: list[tuple[str, Dataset]]
+        data: list[tuple[str, Dataset]],
     ) -> list[tuple[str, Dataset]]:
         result = [
             (bucket[0], bucket[1][column])
@@ -354,7 +351,6 @@ def calc(
         ) = None,
         **kwargs,
     ) -> dict:
-
         if compare_by is None and target_fields_data is None:
             raise ValueError(
                 "You should pass either compare_by or target_fields argument."
@@ -390,9 +386,7 @@ def execute(self, data: ExperimentData) -> ExperimentData:
         )
 
         if len(target_fields_data.columns) == 0:
-            if (
-                data.ds.tmp_roles
-            ):  # if the column is not suitable for the test, then the target will be empty, but if there is a role tempo, then this is normal behavior
+            if data.ds.tmp_roles:  # if the column is not suitable for the test, then the target will be empty, but if there is a role tempo, then this is normal behavior
                 return data
             else:
                 raise NoColumnsError(TargetRole().role_name)

hypex/comparators/abstract.py

@@ -81,7 +81,6 @@ def _inner_function(
 
 
 class PSI(Comparator):
-
     @classmethod
     def _inner_function(
         cls, data: Dataset, test_data: Dataset | None = None, **kwargs

hypex/comparators/comparators.py

@@ -18,7 +18,6 @@
 
 
 class MahalanobisDistance(Calculator):
-
     def __init__(
         self,
         grouping_role: ABCRole | None = None,
@@ -107,7 +106,6 @@ def calc(
             grouping_data, target_fields=target_fields, old_data=data, **kwargs
         )
 
-    # TODO выделить в отдельную функцию с кваргами (нужно для альфы)
     def execute(self, data: ExperimentData) -> ExperimentData:
         group_field, target_fields = self._get_fields(data=data)
         self.key = str(

hypex/comparators/distances.py

@@ -26,7 +26,6 @@ def _inner_function(
 
 
 class KSTest(StatHypothesisTesting):
-
     @property
     def search_types(self) -> list[type] | None:
         return NUMBER_TYPES_LIST
@@ -41,7 +40,6 @@ def _inner_function(
 
 
 class UTest(StatHypothesisTesting):
-
     @property
     def search_types(self) -> list[type] | None:
         return NUMBER_TYPES_LIST
@@ -56,7 +54,6 @@ def _inner_function(
 
 
 class Chi2Test(StatHypothesisTesting):
-
     @property
     def search_types(self) -> list[type] | None:
         return [str]

hypex/comparators/hypothesis_testing.py

@@ -11,7 +11,7 @@
 from .comparators import Comparator
 
 
-class TestPower(Comparator, ABC):
+class PowerTesting(Comparator, ABC):
     def __init__(
         self,
         grouping_role: ABCRole | None = None,
@@ -45,7 +45,7 @@ def execute(self, data: ExperimentData) -> ExperimentData:
         return super().execute(data)
 
 
-class MDEBySize(TestPower):
+class MDEBySize(PowerTesting):
     def _set_value(
         self, data: ExperimentData, value: Dataset | None = None, key: Any = None
     ) -> ExperimentData:

hypex/comparators/test_power.py hypex/comparators/power_testing.pyhypex/comparators/test_power.py renamed to hypex/comparators/power_testing.py

@@ -40,9 +40,7 @@
 
 
 class Dataset(DatasetBase):
-
     class Locker:
-
         def __init__(self, backend, roles):
             self.backend = backend
             self.roles = roles
@@ -348,7 +346,6 @@ def _check_other_dataset(self, other):
     def astype(
         self, dtype: dict[str, type], errors: Literal["raise", "ignore"] = "raise"
     ) -> Dataset:
-
         for col, _ in dtype.items():
             if (errors == "raise") and (col not in self.columns):
                 raise KeyError(f"Column '{col}' does not exist in the Dataset.")
@@ -734,7 +731,6 @@ def replace(
 
 
 class ExperimentData:
-
     def __init__(self, data: Dataset):
         self._data = data
         self.additional_fields = Dataset.create_empty(index=data.index)
@@ -820,7 +816,6 @@ def get_ids(
         searched_space: ExperimentDataEnum | Iterable[ExperimentDataEnum] | None = None,
         key: str | None = None,
     ) -> dict[str, dict[str, list[str]]]:
-
         def check_id(id_: str, class_: str) -> bool:
             result = id_[: id_.find(ID_SPLIT_SYMBOL)] == class_
 
@@ -933,7 +928,6 @@ def field_data_search(
 
 
 class DatasetAdapter(Adapter):
-
     @staticmethod
     def to_dataset(
         data: dict | Dataset | pd.DataFrame | list | str | int | float | bool,

hypex/dataset/dataset.py

@@ -11,7 +11,6 @@
 
 
 class ABCRole(ABC):
-
     _role_name: RoleNameType = "Abstract"
 
     def __init__(self, data_type: DefaultRoleTypes | None = None):
@@ -26,54 +25,46 @@ def __repr__(self) -> str:
 
 
 class InfoRole(ABCRole):
-
     _role_name: RoleNameType = "Info"
 
 
 class StratificationRole(ABCRole):
-
     _role_name: RoleNameType = "Stratification"
 
     def __init__(self, data_type: CategoricalTypes | None = None):
         super().__init__(data_type)
 
 
 class GroupingRole(ABCRole):
-
     _role_name: RoleNameType = "Grouping"
 
     def __init__(self, data_type: CategoricalTypes | None = None):
         super().__init__(data_type)
 
 
 class TreatmentRole(ABCRole):
-
     _role_name: RoleNameType = "Treatment"
 
 
 class TargetRole(ABCRole):
-
     _role_name: RoleNameType = "Target"
 
     def __init__(self, data_type: TargetRoleTypes | None = None):
         super().__init__(data_type)
 
 
 class FeatureRole(ABCRole):
-
     _role_name: RoleNameType = "Feature"
 
 
 class PreTargetRole(ABCRole):
-
     _role_name: RoleNameType = "PreTarget"
 
     def __init__(self, data_type: TargetRoleTypes | None = None):
         super().__init__(data_type)
 
 
 class StatisticRole(ABCRole):
-
     _role_name: RoleNameType = "Statistic"
 
 
@@ -95,58 +86,47 @@ class TempRole(ABCRole):
 
 
 class TempTreatmentRole(TempRole, TreatmentRole):
-
     _role_name: RoleNameType = "TempTreatment"
 
 
 class TempTargetRole(TempRole, TargetRole):
-
     _role_name: RoleNameType = "TempTarget"
 
 
 class TempGroupingRole(TempRole, GroupingRole):
-
     _role_name: RoleNameType = "TempGrouping"
 
 
 class DefaultRole(ABCRole):
-
     _role_name: RoleNameType = "Default"
 
 
 class ReportRole(ABCRole):
-
     _role_name: RoleNameType = "Report"
 
 
 # ___________________________________________________________________________________________
 class AdditionalRole(ABCRole):
-
     _role_name: RoleNameType = "Additional"
 
 
 class AdditionalTreatmentRole(AdditionalRole):
-
     _role_name: RoleNameType = "AdditionalTreatment"
 
 
 class AdditionalGroupingRole(AdditionalRole):
-
     _role_name: RoleNameType = "AdditionalGrouping"
 
 
 class AdditionalTargetRole(AdditionalRole):
-
     _role_name: RoleNameType = "AdditionalTarget"
 
 
 class AdditionalPreTargetRole(AdditionalRole):
-
     _role_name: RoleNameType = "AdditionalPreTarget"
 
 
 class AdditionalMatchingRole(AdditionalRole):
-
     _role_name: RoleNameType = "AdditionalMatching"