fixed some types

pts/evaluation/evaluator.py

@@ -113,7 +113,7 @@ class Evaluator:
     @staticmethod
     def extract_pred_target(
             time_series: Union[pd.Series, pd.DataFrame],
-            forecast: Forecast) -> Union[pd.Series, pd.DataFrame]:
+            forecast: Forecast) -> np.ndarray:
         """
 
         Parameters

pts/modules/distribution_output.py

@@ -68,7 +68,7 @@ class Output(ABC):
         pass
 
 
-class DistributionOutput(Output):
+class DistributionOutput(Output, ABC):
     distr_cls: type
 
     def distribution(

pts/transform/sampler.py

@@ -18,7 +18,7 @@ import numpy as np
 from pts.dataset.stat import ScaleHistogram
 
 
-class InstanceSampler:
+class InstanceSampler(ABC):
     """
     An InstanceSampler is called with the time series and the valid
     index bounds a, b and should return a set of indices a <= i <= b
@@ -41,8 +41,9 @@ class InstanceSampler:
         Selected points to sample
     """
 
+    @abstractmethod
     def __call__(self, ts: np.ndarray, a: int, b: int) -> np.ndarray:
-        raise NotImplementedError()
+        pass
 
 
 class UniformSplitSampler(InstanceSampler):

pts/transform/split.py

@@ -12,7 +12,7 @@
 # permissions and limitations under the License.
 
 from functools import lru_cache
-from typing import Iterator, List, Optional
+from typing import Iterator, List, Optional, Union
 
 import numpy as np
 import pandas as pd
@@ -152,7 +152,7 @@ class InstanceSplitter(FlatMapTransformation):
                 # the prediction length during training, so we just skip these.
                 # If we want to include them we would need to pad and to mask
                 # the loss.
-                sampling_indices: List[int] = []
+                sampling_indices: Union[np.ndarray, List[int]] = []
             else:
                 if self.pick_incomplete:
                     sampling_indices = self.train_sampler(