[SC-7864] Create credit risk scorecard notebook using XGBoost

pyproject.toml

@@ -10,7 +10,7 @@ description = "ValidMind Library"
 license = "Commercial License"
 name = "validmind"
 readme = "README.pypi.md"
-version = "2.7.3"
+version = "2.7.4"
 
 [tool.poetry.dependencies]
 aiohttp = {extras = ["speedups"], version = "*"}

tests/test_integration_tests.py

@@ -44,6 +44,11 @@
     "validmind.data_validation.ProtectedClassesCombination",
     "validmind.data_validation.ProtectedClassesDisparity",
     "validmind.data_validation.ProtectedClassesThresholdOptimizer",
+    # The customer churn classification dataset contains a string column 'Germany' which is not supported by the MutualInformation test
+    "validmind.data_validation.MutualInformation",
+    # The required column 'score' is not present in the dataset with input_id test_dataset
+    "validmind.data_validation.ScoreBandDefaultRates",
+    "validmind.model_validation.sklearn.ScoreProbabilityAlignment",
 ]
 SKIPPED_TESTS = []
 SUCCESSFUL_TESTS = []

tests/unit_tests/data_validation/test_TooManyZeroValues.py

@@ -37,7 +37,7 @@ def setUp(self):
         )
 
     def test_too_many_zeros_default_threshold(self):
-        # Test with default threshold (3%)
+        # Test with default threshold (0.03 or 0.03%)
         results, passed = TooManyZeroValues(self.vm_dataset)
 
         # Check return types
@@ -47,22 +47,25 @@ def test_too_many_zeros_default_threshold(self):
         # Check results structure
         self.assertEqual(len(results), 2)  # Should only check numeric columns
         for result in results:
-            self.assertIn("Column", result)
+            self.assertIn("Variable", result)
+            self.assertIn("Row Count", result)
             self.assertIn("Number of Zero Values", result)
             self.assertIn("Percentage of Zero Values (%)", result)
             self.assertIn("Pass/Fail", result)
 
         # Verify specific results
-        few_zeros_result = next(r for r in results if r["Column"] == "few_zeros")
-        many_zeros_result = next(r for r in results if r["Column"] == "many_zeros")
+        few_zeros_result = next(r for r in results if r["Variable"] == "few_zeros")
+        many_zeros_result = next(r for r in results if r["Variable"] == "many_zeros")
 
-        self.assertEqual(few_zeros_result["Pass/Fail"], "Pass")  # 1% should pass
-        self.assertEqual(many_zeros_result["Pass/Fail"], "Fail")  # 5% should fail
-        self.assertFalse(passed)  # Overall test should fail due to many_zeros
+        # 1% should fail as it's > 0.03%
+        self.assertEqual(few_zeros_result["Pass/Fail"], "Fail")
+        # 5% should fail as it's > 0.03%
+        self.assertEqual(many_zeros_result["Pass/Fail"], "Fail")
+        self.assertFalse(passed)  # Overall test should fail as both columns fail
 
     def test_custom_threshold(self):
         # Test with higher threshold (6%)
-        results, passed = TooManyZeroValues(self.vm_dataset, max_percent_threshold=0.06)
+        results, passed = TooManyZeroValues(self.vm_dataset, max_percent_threshold=6)
 
         # Both columns should pass with higher threshold
         self.assertTrue(passed)

tests/unit_tests/data_validation/test_UniqueRows.py

@@ -55,7 +55,10 @@ def test_unique_rows_default_threshold(self):
 
     def test_low_uniqueness(self):
         # Test dataset with low uniqueness
-        results, passed = UniqueRows(self.vm_dataset_duplicates)
+        # Set threshold to 5% to make the test fail
+        results, passed = UniqueRows(
+            self.vm_dataset_duplicates, min_percent_threshold=5
+        )
 
         # Check return types
         self.assertIsInstance(results, list)
@@ -68,6 +71,6 @@ def test_low_uniqueness(self):
             self.assertIn("Percentage of Unique Values (%)", result)
             self.assertIn("Pass/Fail", result)
 
-        # Should fail with low uniqueness
+        # Should fail with uniqueness below 5%
         self.assertFalse(passed)
-        self.assertTrue(any(row["Pass/Fail"] == "Fail" for row in results))
+        self.assertTrue(all(row["Pass/Fail"] == "Fail" for row in results))

validmind/__version__.py

@@ -1 +1 @@
-__version__ = "2.7.3"
+__version__ = "2.7.4"