binboost 0.2.3__tar.gz → 0.2.4__tar.gz

{binboost-0.2.3 → binboost-0.2.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: binboost
-Version: 0.2.3
+Version: 0.2.4
 Summary: Gradient Boosting Berbasis Aturan Logika Adaptif untuk Fitur Biner
 Author: Rangga Wahyu Pratama
 License: MIT

{binboost-0.2.3 → binboost-0.2.4}/binboost/__init__.py

@@ -1,6 +1,6 @@
 from .binboost import BinBoost
 from .multiclass import OneVsRestBinBoost
 
-__version__ = "0.2.3"
+__version__ = "0.2.4"
 __author__ = "Rangga Wahyu Pratama"
-__all__ = ["BinBoost"]
+__all__ = ["BinBoost", "OneVsRestBinBoost"]

{binboost-0.2.3 → binboost-0.2.4}/binboost/binarizer.py

@@ -4,7 +4,6 @@ import numpy as np
 
 class GradientBinarizer:
     # Binarisasi fitur numerik berbasis korelasi dengan gradien saat ini
-
     def __init__(self, strategy='gradient', n_thresholds='auto'):
         self.strategy = strategy
         self.n_thresholds = n_thresholds

{binboost-0.2.3 → binboost-0.2.4}/binboost/binboost.py

@@ -309,6 +309,7 @@ class BinBoost:
         )
 
         numeric_cols = self._detect_numeric_cols(X)
+        self.numeric_cols_ = numeric_cols
         self.ohe_groups_ = self._detect_ohe_groups(X)
 
         self._binarizer = GradientBinarizer(
@@ -454,7 +455,7 @@ class BinBoost:
         self._check_is_fitted()
         X, _ = self._validate_input(X)
         F = np.zeros(X.shape[0])
-        numeric_cols = self._detect_numeric_cols(X)
+        numeric_cols = self.numeric_cols_
         for i, (rule, weights) in enumerate(zip(self.estimators_, self.rule_weights_)):
             if self.binarize_strategy == 'gradient' and numeric_cols:
                 X_bin = self._binarize_untuk_iterasi(X, i, numeric_cols)
@@ -521,7 +522,7 @@ class BinBoost:
         """
         self._check_is_fitted()
         X, _ = self._validate_input(X)
-        numeric_cols = self._detect_numeric_cols(X)
+        numeric_cols = self.numeric_cols_
         F = np.zeros(X.shape[0])
         for i, (rule, weights) in enumerate(zip(self.estimators_, self.rule_weights_)):
             if self.binarize_strategy == 'gradient' and numeric_cols:
@@ -556,7 +557,7 @@ class BinBoost:
         """
         self._check_is_fitted()
         X, _ = self._validate_input(X)
-        numeric_cols = self._detect_numeric_cols(X)
+        numeric_cols = self.numeric_cols_
         results = []
         
         for i, rule in enumerate(self.estimators_):
@@ -565,8 +566,8 @@ class BinBoost:
             else:
                 dummy_grad = np.ones(X.shape[0])
                 X_bin, _ = self._binarizer.transform(X, dummy_grad, numeric_cols)
-                results.append(rule.evaluate(X_bin))
-                
+            results.append(rule.evaluate(X_bin))
+            
         return np.column_stack(results)
 
     def get_params(self, deep=True):

{binboost-0.2.3 → binboost-0.2.4}/binboost/multiclass.py

@@ -4,12 +4,6 @@ from .binboost import BinBoost
 
 
 class OneVsRestBinBoost:
-    """
-    Ekstensi BinBoost untuk klasifikasi multi-kelas (>2 kelas) lewat strategi
-    One-vs-Rest, mirip pola sklearn.multiclass.OneVsRestClassifier.
-    Untuk 2 kelas, gunakan langsung BinBoost.
-    """
-
     def __init__(self, mode='multiclass', **binboost_params):
         self.mode = mode
         self.binboost_params = binboost_params

{binboost-0.2.3 → binboost-0.2.4}/binboost.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: binboost
-Version: 0.2.3
+Version: 0.2.4
 Summary: Gradient Boosting Berbasis Aturan Logika Adaptif untuk Fitur Biner
 Author: Rangga Wahyu Pratama
 License: MIT

{binboost-0.2.3 → binboost-0.2.4}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "binboost"
-version = "0.2.3"
+version = "0.2.4"
 description = "Gradient Boosting Berbasis Aturan Logika Adaptif untuk Fitur Biner"
 readme = "README.md"
 requires-python = ">=3.8"

{binboost-0.2.3 → binboost-0.2.4}/setup.py

@@ -2,7 +2,7 @@ from setuptools import setup, find_packages
 
 setup(
     name="binboost",
-    version="0.2.3",
+    version="0.2.4",
     author="Rangga Wahyu Pratama",
     description="Gradient Boosting Berbasis Aturan Logika Adaptif untuk Fitur Biner",
     long_description=open("README.md", encoding="utf-8").read(),