Fix: Handle bracket-wrapped parameter values in XGBoost config

- Fixed extract_model_param() to handle values like '[9.9E-2]'
- Added comprehensive tests for both bracket and non-bracket formats
- Bumped version to 0.2.6.1 (maintenance release per PEP 440)
- Fixes #5

Author: xRiskLab

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "xbooster"
-version = "0.2.6.post1"
+version = "0.2.6.1"
 description = "Explainable Boosted Scoring"
 authors = [
     {name = "xRiskLab", email = "contact@xrisklab.ai"}

tests/test_extract_model_param_formats.py

@@ -0,0 +1,138 @@
+"""
+Test module for extract_model_param with different value formats.
+
+This test ensures that the extract_model_param method can handle different
+string formats returned by XGBoost config, including values with brackets.
+
+GitHub Issue: Value of base_score parameter is read as a string wrapped in
+list brackets (e.g., '[9.9E-2]'), which caused a ValueError when attempting
+to convert it directly to a float.
+"""
+
+import json
+from unittest.mock import patch
+
+import pandas as pd
+import pytest
+import xgboost as xgb
+
+from xbooster.xgb_constructor import XGBScorecardConstructor
+
+
+@pytest.fixture(scope="module")
+def simple_model():
+    """
+    Creates and trains a simple XGBoost model.
+
+    Returns:
+        model (xgb.XGBClassifier): Trained XGBoost model.
+    """
+    X = pd.DataFrame({"feature1": [1, 2, 3, 4, 5], "feature2": [5, 4, 3, 2, 1]})
+    y = pd.Series([0, 1, 0, 1, 0])
+    model = xgb.XGBClassifier(n_estimators=5, max_depth=1, random_state=42)
+    model.fit(X, y)
+    return model
+
+
+@pytest.fixture(scope="module")
+def simple_data():
+    """
+    Creates simple training data.
+
+    Returns:
+        tuple: X and y DataFrames.
+    """
+    X = pd.DataFrame({"feature1": [1, 2, 3, 4, 5], "feature2": [5, 4, 3, 2, 1]})
+    y = pd.Series([0, 1, 0, 1, 0])
+    return X, y
+
+
+def test_extract_model_param_with_brackets(simple_model, simple_data):
+    """
+    Test that extract_model_param handles values with brackets correctly.
+
+    This test verifies the fix for GitHub issue where base_score came as '[9.9E-2]'
+    instead of '9.9E-2', causing a ValueError.
+    """
+    X, y = simple_data
+    constructor = XGBScorecardConstructor(simple_model, X, y)
+
+    # Test with bracket format
+    mock_config = {
+        "learner": {
+            "learner_model_param": {"base_score": "[9.9E-2]"},
+            "gradient_booster": {
+                "tree_train_param": {"learning_rate": "[0.3]", "max_depth": "[1]"}
+            },
+        }
+    }
+
+    with patch.object(constructor.booster_, "save_config", return_value=json.dumps(mock_config)):
+        base_score = constructor.extract_model_param("base_score")
+        learning_rate = constructor.extract_model_param("learning_rate")
+        max_depth = constructor.extract_model_param("max_depth")
+
+    assert abs(base_score - 0.099) < 1e-6
+    assert abs(learning_rate - 0.3) < 1e-6
+    assert abs(max_depth - 1.0) < 1e-6
+
+
+def test_extract_model_param_without_brackets(simple_model, simple_data):
+    """
+    Test that extract_model_param handles values without brackets correctly.
+
+    This ensures backward compatibility with the standard format.
+    """
+    X, y = simple_data
+    constructor = XGBScorecardConstructor(simple_model, X, y)
+
+    # Test without bracket format (standard)
+    mock_config = {
+        "learner": {
+            "learner_model_param": {"base_score": "9.9E-2"},
+            "gradient_booster": {"tree_train_param": {"learning_rate": "0.3", "max_depth": "1"}},
+        }
+    }
+
+    with patch.object(constructor.booster_, "save_config", return_value=json.dumps(mock_config)):
+        base_score = constructor.extract_model_param("base_score")
+        learning_rate = constructor.extract_model_param("learning_rate")
+        max_depth = constructor.extract_model_param("max_depth")
+
+    assert abs(base_score - 0.099) < 1e-6
+    assert abs(learning_rate - 0.3) < 1e-6
+    assert abs(max_depth - 1.0) < 1e-6
+
+
+def test_extract_model_param_various_formats(simple_model, simple_data):
+    """
+    Test that extract_model_param handles various numeric string formats.
+    """
+    X, y = simple_data
+    constructor = XGBScorecardConstructor(simple_model, X, y)
+
+    test_cases = [
+        ("5E-1", 0.5),
+        ("[5E-1]", 0.5),
+        ("1.0", 1.0),
+        ("[1.0]", 1.0),
+        ("0.099", 0.099),
+        ("[0.099]", 0.099),
+    ]
+
+    for value_str, expected in test_cases:
+        mock_config = {
+            "learner": {
+                "learner_model_param": {"base_score": value_str},
+                "gradient_booster": {"tree_train_param": {"learning_rate": "0.3"}},
+            }
+        }
+
+        with patch.object(
+            constructor.booster_, "save_config", return_value=json.dumps(mock_config)
+        ):
+            result = constructor.extract_model_param("base_score")
+
+        assert abs(result - expected) < 1e-6, (
+            f"Failed for {value_str}: got {result}, expected {expected}"
+        )

uv.lock

@@ -109,8 +109,16 @@ def extract_model_param(self, param):
         """
         config = json.loads(self.booster_.save_config())
         if param == "base_score":
-            return float(config["learner"]["learner_model_param"][param])
-        return float(config["learner"]["gradient_booster"]["tree_train_param"][param])
+            value = config["learner"]["learner_model_param"][param]
+        else:
+            value = config["learner"]["gradient_booster"]["tree_train_param"][param]
+
+        # Handle different value formats from XGBoost config
+        if isinstance(value, str):
+            # Remove brackets if present (e.g., '[9.9E-2]' -> '9.9E-2')
+            value = value.strip("[]")
+
+        return float(value)
 
     def add_detailed_split(self, dataframe: Optional[pd.DataFrame] = None) -> pd.DataFrame:
         """