Fix tests failing due to scipy 1.14 (#601)

Author: lajohn4747

pyproject.toml

@@ -30,9 +30,9 @@ dependencies = [
     "scikit-learn>=1.1.0;python_version>='3.10' and python_version<'3.11'",
     "scikit-learn>=1.1.3;python_version>='3.11' and python_version<'3.12'",
     "scikit-learn>=1.3.1;python_version>='3.12'",
-    "scipy>=1.7.3,<1.14.0;python_version<'3.10'",
-    "scipy>=1.9.2,<1.14.0;python_version>='3.10' and python_version<'3.12'",
-    "scipy>=1.12.0,<1.14.0;python_version>='3.12'",
+    "scipy>=1.7.3;python_version<'3.10'",
+    "scipy>=1.9.2;python_version>='3.10' and python_version<'3.12'",
+    "scipy>=1.12.0;python_version>='3.12'",
     'copulas>=0.11.0',
     'tqdm>=4.29',
     'plotly>=5.19.0',

tests/integration/reports/multi_table/test_quality_report.py

@@ -107,7 +107,7 @@ def test_multi_table_quality_report():
         details.append(report.get_details(property_))
 
     # Assert score
-    assert score == 0.649582127409184
+    assert round(score, 15) == 0.649582127409184
     pd.testing.assert_frame_equal(
         properties,
         pd.DataFrame({

tests/integration/reports/single_table/_properties/test_column_pair_trends.py

@@ -1,5 +1,6 @@
 import numpy as np
 import pandas as pd
+from tests.utils import get_error_type
 
 from sdmetrics.demos import load_demo
 from sdmetrics.reports.single_table._properties.column_pair_trends import ColumnPairTrends
@@ -79,17 +80,25 @@ def test_get_score_warnings(self, recwarn):
         # Run
         column_pair_trends = ColumnPairTrends()
 
-        exp_message_1 = "ValueError: could not convert string to float: 'a'"
+        exp_message_1 = 'ValueError'
 
-        exp_message_2 = "TypeError: '<=' not supported between instances of 'float' and 'str'"
+        exp_message_2 = 'TypeError'
 
-        exp_error_serie = pd.Series([exp_message_1, None, None, exp_message_2, exp_message_2, None])
+        exp_error_series = pd.Series([
+            exp_message_1,
+            None,
+            None,
+            exp_message_2,
+            exp_message_2,
+            None,
+        ])
 
         score = column_pair_trends.get_score(real_data, synthetic_data, metadata)
 
         # Assert
         details = column_pair_trends.details
-        pd.testing.assert_series_equal(details['Error'], exp_error_serie, check_names=False)
+        details['Error'] = details['Error'].apply(get_error_type)
+        pd.testing.assert_series_equal(details['Error'], exp_error_series, check_names=False)
         assert score == 0.7751937984496124
 
     def test_only_categorical_columns(self):

tests/integration/reports/single_table/test_quality_report.py

@@ -6,6 +6,7 @@
 
 from sdmetrics.demos import load_demo
 from sdmetrics.reports.single_table import QualityReport
+from tests.utils import get_error_type
 
 
 class TestQualityReport:
@@ -262,7 +263,7 @@ def test_report_end_to_end_with_errors(self):
             'Score': [0.6621621621621622, np.nan, 0.9953488372093023, 0.9395348837209302],
             'Error': [
                 None,
-                "TypeError: '<' not supported between instances of 'str' and 'float'",
+                'TypeError',
                 None,
                 None,
             ],
@@ -304,23 +305,25 @@ def test_report_end_to_end_with_errors(self):
             'Real Correlation': [np.nan] * 6,
             'Synthetic Correlation': [np.nan] * 6,
             'Error': [
-                "ValueError: could not convert string to float: 'a'",
+                'ValueError',
                 None,
                 None,
-                "TypeError: '<=' not supported between instances of 'float' and 'str'",
-                "TypeError: '<=' not supported between instances of 'float' and 'str'",
+                'TypeError',
+                'TypeError',
                 None,
             ],
         }
         expected_details_column_shapes = pd.DataFrame(expected_details_column_shapes_dict)
         expected_details_cpt = pd.DataFrame(expected_details_cpt__dict)
 
-        pd.testing.assert_frame_equal(
-            report.get_details('Column Shapes'), expected_details_column_shapes
-        )
-        pd.testing.assert_frame_equal(
-            report.get_details('Column Pair Trends'), expected_details_cpt
-        )
+        # Errors may change based on versions of scipy installed.
+        col_shape_report = report.get_details('Column Shapes')
+        col_pair_report = report.get_details('Column Pair Trends')
+        col_shape_report['Error'] = col_shape_report['Error'].apply(get_error_type)
+        col_pair_report['Error'] = col_pair_report['Error'].apply(get_error_type)
+
+        pd.testing.assert_frame_equal(col_shape_report, expected_details_column_shapes)
+        pd.testing.assert_frame_equal(col_pair_report, expected_details_cpt)
         assert report.get_score() == 0.8204378797402054
 
     def test_report_with_column_nan(self):
@@ -446,10 +449,10 @@ def test_report_with_column_nan(self):
                 None,
                 None,
                 None,
-                'ValueError: x and y must have length at least 2.',
+                'ValueError',
                 None,
                 None,
-                'ValueError: x and y must have length at least 2.',
+                'ValueError',
                 None,
                 None,
                 None,
@@ -458,12 +461,17 @@ def test_report_with_column_nan(self):
         expected_details_column_shapes = pd.DataFrame(expected_details_column_shapes_dict)
         expected_details_cpt = pd.DataFrame(expected_details_cpt__dict)
 
-        pd.testing.assert_frame_equal(
-            report.get_details('Column Shapes'), expected_details_column_shapes
-        )
-        pd.testing.assert_frame_equal(
-            report.get_details('Column Pair Trends'), expected_details_cpt
-        )
+        col_shape_report = report.get_details('Column Shapes')
+        if 'Error' not in col_shape_report:
+            # Errors may not occur in certain scipy versions
+            expected_details_column_shapes.drop(columns=['Error'], inplace=True)
+
+        # Errors may change based on versions of library installed.
+        col_pair_report = report.get_details('Column Pair Trends')
+        col_pair_report['Error'] = col_pair_report['Error'].apply(get_error_type)
+
+        pd.testing.assert_frame_equal(col_shape_report, expected_details_column_shapes)
+        pd.testing.assert_frame_equal(col_pair_report, expected_details_cpt)
 
     def test_report_with_verbose(self, capsys):
         """Test the report with verbose.

tests/utils.py

@@ -40,3 +40,10 @@ def __eq__(self, other):
         """Assert equality by expanding the iterator."""
         assert all(x == y for x, y in zip(self.iterator, other))
         return True
+
+
+def get_error_type(error):
+    if error is not None:
+        colon_index = error.find(':')
+        return error[:colon_index]
+    return None