proscore 0.2.2__tar.gz → 0.2.3__tar.gz

{proscore-0.2.2/src/proscore.egg-info → proscore-0.2.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: proscore
-Version: 0.2.2
+Version: 0.2.3
 Summary: Production-grade scorecard development toolkit
 Author: Liqiwei
 License-Expression: MIT

{proscore-0.2.2 → proscore-0.2.3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "proscore"
-version = "0.2.2"
+version = "0.2.3"
 description = "Production-grade scorecard development toolkit"
 readme = "README.md"
 license = "MIT"

{proscore-0.2.2 → proscore-0.2.3}/src/proscore/__init__.py

@@ -19,7 +19,7 @@ from proscore.rules import RuleMiner
 from proscore.selection import Filter, StepwiseSelector, assess_screen
 from proscore.transform import WOETransformer
 
-__version__ = "0.2.2"
+__version__ = "0.2.3"
 
 
 class ProScore:
@@ -104,15 +104,13 @@ class ProScore:
 
     def _categorical_features(self) -> list[str]:
         return [
-            c for c in self.train_df.columns
-            if c != self.target and not pd.api.types.is_numeric_dtype(self.train_df[c])
+            c for c in self.train_df.columns if c != self.target and not pd.api.types.is_numeric_dtype(self.train_df[c])
         ]
 
     def _initial_features(self) -> list[str]:
         """Numeric columns — the Filter candidate pool."""
         return [
-            c for c in self.train_df.columns
-            if c != self.target and pd.api.types.is_numeric_dtype(self.train_df[c])
+            c for c in self.train_df.columns if c != self.target and pd.api.types.is_numeric_dtype(self.train_df[c])
         ]
 
     def _numeric_after_prefilter(self) -> list[str]:
@@ -199,7 +197,8 @@ class ProScore:
         features = self._initial_features()
         self._prefilter = Filter(**kwargs)
         self._prefilter.fit(
-            self._train_X(features), self._train_y(),
+            self._train_X(features),
+            self._train_y(),
             X_test=self._test_X(features) if self.test_df is not None else None,
             bin_table=None,
         )
@@ -278,7 +277,8 @@ class ProScore:
         self._refine_skipped = False
         self._filter = Filter(**kwargs)
         self._filter.fit(
-            self._train_X(features), self._train_y(),
+            self._train_X(features),
+            self._train_y(),
             X_test=self._test_X(features) if self.test_df is not None else None,
             bin_table=self._binner.bin_table_,
         )
@@ -358,10 +358,8 @@ class ProScore:
             _warn_halted(self)
             return self
         _check_transformer(self)
-        features = [c for c in self._features_for_modeling()
-                    if c not in self._rule_features]
+        features = [c for c in self._features_for_modeling() if c not in self._rule_features]
         train_woe = self._transformer.transform(self._train_X(features))
-        train_woe[self.target] = self._train_y().values
 
         test_woe = None
         y_test = None
@@ -372,8 +370,12 @@ class ProScore:
         force_in = kwargs.pop("force_in", None)
         self._selector = StepwiseSelector(**kwargs)
         self._selector.fit(
-            train_woe, self._train_y(), candidates=features,
-            force_in=force_in, X_test=test_woe, y_test=y_test,
+            train_woe,
+            self._train_y(),
+            candidates=features,
+            force_in=force_in,
+            X_test=test_woe,
+            y_test=y_test,
         )
         return self
 
@@ -419,10 +421,14 @@ class ProScore:
 
         self.eval_result = _evaluate(
             self._scorecard.model_,
-            train_woe, self._train_y(),
-            X_test=test_woe, y_test=self._test_y(),
-            X_oot=oot_woe, y_oot=self._oot_y(),
-            features=features, n_bins=n_bins,
+            train_woe,
+            self._train_y(),
+            X_test=test_woe,
+            y_test=self._test_y(),
+            X_oot=oot_woe,
+            y_oot=self._oot_y(),
+            features=features,
+            n_bins=n_bins,
         )
         return self
 

{proscore-0.2.2 → proscore-0.2.3}/src/proscore/__main__.py

@@ -37,8 +37,7 @@ def _cmd_run(args: list[str]) -> None:
 
     parser = argparse.ArgumentParser(description="Run a pipeline from Excel config")
     parser.add_argument("config", help="Path to pipeline.xlsx")
-    parser.add_argument("--output-script", "-o", default=None,
-                        help="Also generate a self-contained Python script")
+    parser.add_argument("--output-script", "-o", default=None, help="Also generate a self-contained Python script")
     opts = parser.parse_args(args)
 
     from proscore._pipeline_config import run_pipeline
@@ -51,6 +50,7 @@ def _cmd_run(args: list[str]) -> None:
     except Exception as e:
         print(f"运行错误: {e}", file=sys.stderr)
         import traceback
+
         traceback.print_exc()
         sys.exit(1)
 

{proscore-0.2.2 → proscore-0.2.3}/src/proscore/_data/__init__.py

@@ -62,13 +62,15 @@ class DataReader:
         rows = []
         for col in self.features_:
             s = self.df[col]
-            rows.append({
-                "variable": col,
-                "dtype": str(s.dtype),
-                "n_missing": int(s.isna().sum()),
-                "missing_pct": round(s.isna().mean() * 100, 2),
-                "n_unique": int(s.nunique()),
-            })
+            rows.append(
+                {
+                    "variable": col,
+                    "dtype": str(s.dtype),
+                    "n_missing": int(s.isna().sum()),
+                    "missing_pct": round(s.isna().mean() * 100, 2),
+                    "n_unique": int(s.nunique()),
+                }
+            )
         return pd.DataFrame(rows).sort_values("missing_pct", ascending=False).reset_index(drop=True)
 
     def __repr__(self) -> str: