use replace()

R-Palazzo · R-Palazzo · commit 568b7227d212 · 2025-03-10T10:44:00.000-04:00
diff --git a/sdmetrics/single_table/bayesian_network.py b/sdmetrics/single_table/bayesian_network.py
@@ -44,11 +44,8 @@ def _likelihoods(cls, real_data, synthetic_data, metadata=None, structure=None):
             }
             for column in fields
         }
-        for column in fields:
-            real_data[column] = real_data[column].map(category_to_integer[column]).astype('int64')
-            synthetic_data[column] = (
-                synthetic_data[column].map(category_to_integer[column]).astype(int)
-            )
+        real_data[fields] = real_data[fields].replace(category_to_integer).astype('int64')
+        synthetic_data[fields] = synthetic_data[fields].replace(category_to_integer).astype('int64')
 
         bn.fit(torch.tensor(real_data[fields].to_numpy()))
         LOGGER.debug('Evaluating likelihood of the synthetic data')

Original file line number	Diff line number	Diff line change
`@@ -44,11 +44,8 @@ def _likelihoods(cls, real_data, synthetic_data, metadata=None, structure=None):`
`44`	`44`	`}`
`45`	`45`	`for column in fields`
`46`	`46`	`}`
`47`		`- for column in fields:`
`48`		`- real_data[column] = real_data[column].map(category_to_integer[column]).astype('int64')`
`49`		`- synthetic_data[column] = (`
`50`		`- synthetic_data[column].map(category_to_integer[column]).astype(int)`
`51`		`- )`
	`47`	`+ real_data[fields] = real_data[fields].replace(category_to_integer).astype('int64')`
	`48`	`+ synthetic_data[fields] = synthetic_data[fields].replace(category_to_integer).astype('int64')`
`52`	`49`
`53`	`50`	`bn.fit(torch.tensor(real_data[fields].to_numpy()))`
`54`	`51`	`LOGGER.debug('Evaluating likelihood of the synthetic data')`