Consistent string formatting.

PiperOrigin-RevId: 326007570

tensorflow_privacy/privacy/membership_inference_attack/data_structures.py

@@ -56,7 +56,7 @@ class SingleSliceSpec:
     if self.feature == SlicingFeature.PERCENTILE:
       return 'Loss percentiles: %d-%d' % self.value
 
-    return f'{self.feature.name}={self.value}'
+    return '%s=%s' % (self.feature.name, self.value)
 
 
 @dataclass
@@ -98,7 +98,7 @@ class AttackType(enum.Enum):
 
   # Return LOGISTIC_REGRESSION instead of AttackType.LOGISTIC_REGRESSION
   def __str__(self):
-    return f'{self.name}'
+    return '%s' % self.name
 
 
 @dataclass
@@ -125,7 +125,7 @@ class AttackInputData:
   def num_classes(self):
     if self.labels_train is None or self.labels_test is None:
       raise ValueError(
-          "Can't identify the number of classes as no labels were provided. "
+          'Can\'t identify the number of classes as no labels were provided. '
           'Please set labels_train and labels_test')
     return int(max(np.max(self.labels_train), np.max(self.labels_test))) + 1
 
@@ -204,7 +204,7 @@ class RocCurve:
     over all available classifier thresholds.
 
     Returns:
-      a single float number with membership attaker's advantage.
+      a single float number with membership attacker's advantage.
     """
     return max(np.abs(self.tpr - self.fpr))
 

tensorflow_privacy/privacy/membership_inference_attack/dataset_slicing.py

@@ -137,7 +137,7 @@ def get_slice(data: AttackInputData,
   elif slice_spec.feature == SlicingFeature.CORRECTLY_CLASSIFIED:
     data_slice = _slice_by_classification_correctness(data, slice_spec.value)
   else:
-    raise ValueError(f'Unknown slice spec feature "{slice_spec.feature}"')
+    raise ValueError('Unknown slice spec feature "%s"' % slice_spec.feature)
 
   data_slice.slice_spec = slice_spec
   return data_slice

tensorflow_privacy/privacy/membership_inference_attack/membership_inference_attack_new.py

@@ -54,8 +54,8 @@ def run_trained_attack(attack_input: AttackInputData, attack_type: AttackType):
   elif attack_type == AttackType.K_NEAREST_NEIGHBORS:
     attacker = models.KNearestNeighborsAttacker()
   else:
-    raise NotImplementedError(
-        'Attack type {} not implemented yet.'.format(attack_type))
+    raise NotImplementedError('Attack type %s not implemented yet.' %
+                              attack_type)
 
   prepared_attacker_data = models.create_attacker_data(attack_input)