Try to improve performance of contingency_similarity

pyproject.toml

@@ -186,6 +186,7 @@ exclude = [
     ".git",
     "__pycache__",
     ".ipynb_checkpoints",
+    ".ipynb",
     "tasks.py",
 ]
 

sdmetrics/column_pairs/statistical/contingency_similarity.py

@@ -1,7 +1,5 @@
 """Contingency Similarity Metric."""
 
-import pandas as pd
-
 from sdmetrics.column_pairs.base import ColumnPairsMetric
 from sdmetrics.goal import Goal
 
@@ -42,43 +40,16 @@ def compute(cls, real_data, synthetic_data):
         columns = real_data.columns[:2]
         real = real_data[columns]
         synthetic = synthetic_data[columns]
-        contingency_real = pd.crosstab(
-            index=real[columns[0]].astype(str),
-            columns=real[columns[1]].astype(str),
-            normalize=True,
-        )
-        contingency_synthetic = pd.crosstab(
-            index=synthetic[columns[0]].astype(str),
-            columns=synthetic[columns[1]].astype(str),
-            normalize=True,
-        )
-
-        real_append_cols = {}
-        for col in set(contingency_synthetic.columns) - set(contingency_real.columns):
-            real_append_cols[col] = [0 for _ in range(len(contingency_real))]
-        contingency_real = pd.concat(
-            (contingency_real, pd.DataFrame(real_append_cols, index=contingency_real.index)),
-            axis=1,
+        contingency_real = real.groupby(list(columns), dropna=False).size() / len(real)
+        contingency_synthetic = synthetic.groupby(list(columns), dropna=False).size() / len(
+            synthetic
         )
-
-        synthetic_append_cols = {}
-        for col in set(contingency_real.columns) - set(contingency_synthetic.columns):
-            synthetic_append_cols[col] = [0 for _ in range(len(contingency_synthetic))]
-        contingency_synthetic = pd.concat(
-            (
-                contingency_synthetic,
-                pd.DataFrame(synthetic_append_cols, index=contingency_synthetic.index),
-            ),
-            axis=1,
-        )
-
-        for row in set(contingency_synthetic.index) - set(contingency_real.index):
-            contingency_real.loc[row] = [0 for _ in range(len(contingency_real.columns))]
-        for row in set(contingency_real.index) - set(contingency_synthetic.index):
-            contingency_synthetic.loc[row] = [0 for _ in range(len(contingency_synthetic.columns))]
-
-        variation = abs(contingency_real - contingency_synthetic) / 2
-        return 1 - variation.sum().sum()
+        combined_index = contingency_real.index.union(contingency_synthetic.index)
+        contingency_synthetic = contingency_synthetic.reindex(combined_index, fill_value=0)
+        contingency_real = contingency_real.reindex(combined_index, fill_value=0)
+        diff = abs(contingency_real - contingency_synthetic).fillna(0)
+        variation = diff / 2
+        return 1 - variation.sum()
 
     @classmethod
     def normalize(cls, raw_score):

tests/integration/reports/multi_table/_properties/test_inter_table_trends.py

@@ -17,7 +17,7 @@ def test_end_to_end(self):
         result = inter_table_trends.get_score(real_data, synthetic_data, metadata)
 
         # Assert
-        assert result == 0.4416666666666667
+        assert result == 0.4416666666666666
 
     def test_with_progress_bar(self):
         """Test that the progress bar is correctly updated."""
@@ -38,5 +38,5 @@ def test_with_progress_bar(self):
         result = inter_table_trends.get_score(real_data, synthetic_data, metadata, progress_bar)
 
         # Assert
-        assert result == 0.4416666666666667
+        assert result == 0.4416666666666666
         assert mock_update.call_count == num_iter

tests/integration/reports/multi_table/test_quality_report.py

@@ -232,9 +232,9 @@ def test_quality_report_end_to_end():
     # Assert
     expected_properties = pd.DataFrame({
         'Property': ['Column Shapes', 'Column Pair Trends', 'Cardinality', 'Intertable Trends'],
-        'Score': [0.7978174603174604, 0.45654629583521095, 0.95, 0.4416666666666667],
+        'Score': [0.7978174603174604, 0.45654629583521095, 0.95, 0.4416666666666666],
     })
-    assert score == 0.6615076057048345
+    assert score == 0.6615076057048344
     pd.testing.assert_frame_equal(properties, expected_properties)
     expected_info_keys = {
         'report_type',
@@ -272,9 +272,9 @@ def test_quality_report_with_object_datetimes():
     # Assert
     expected_properties = pd.DataFrame({
         'Property': ['Column Shapes', 'Column Pair Trends', 'Cardinality', 'Intertable Trends'],
-        'Score': [0.7978174603174604, 0.45654629583521095, 0.95, 0.4416666666666667],
+        'Score': [0.7978174603174604, 0.45654629583521095, 0.95, 0.4416666666666666],
     })
-    assert score == 0.6615076057048345
+    assert score == 0.6615076057048344
     pd.testing.assert_frame_equal(properties, expected_properties)
 
 
@@ -342,7 +342,7 @@ def test_quality_report_with_errors():
             None,
         ],
     })
-    assert score == 0.7249603174603175
+    assert score == 0.7249603174603174
     pd.testing.assert_frame_equal(properties, expected_properties)
     pd.testing.assert_frame_equal(details_column_shapes, expected_details)