feat: more informative errors

src/timeseriesflattener/flattened_ds_validator.py

@@ -33,16 +33,18 @@ def _check_for_duplicate_rows(self):
             col_subset=[self.id_col_name, self.timestamp_col_name],
         ):
             raise ValueError(
-                "Duplicate patient/timestamp combinations in prediction_times_df, aborting",
+                "Duplicate id/timestamp combinations in prediction_times_df, aborting",
             )
 
     def _check_that_timestamp_and_id_columns_exist(self):
         """Check that the required columns are present in the initial
-        dataframe."""
+
+        dataframe.
+        """
 
         for col_name in (self.timestamp_col_name, self.id_col_name):
             if col_name not in self.df.columns:
-                raise ValueError(
+                raise KeyError(
                     f"{col_name} does not exist in prediction_times_df, change the df or set another argument",
                 )
 

tests/test_timeseriesflattener/test_flattened_dataset/test_errors.py

@@ -18,7 +18,7 @@ def test_col_does_not_exist_in_prediction_times():
 
     prediction_times_df = str_to_df(prediction_times_str)
 
-    with pytest.raises(ValueError):
+    with pytest.raises(KeyError, match=r".*does not exist.*"):
         TimeseriesFlattener(  # noqa
             prediction_times_df=prediction_times_df,
             timestamp_col_name="timestamp",
@@ -60,10 +60,10 @@ def test_col_does_not_exist():
 
 
 def test_duplicate_prediction_times():
-    with pytest.raises(ValueError):
+    with pytest.raises(ValueError, match=r".*duplicate.*"):
         prediction_times_df_str = """dw_ek_borger,timestamp,
-                                1,2021-12-31 00:00:00
-                                1,2021-11-31 00:00:00
+                                1,2021-12-30 00:00:00
+                                1,2021-12-30 00:00:00
                                 """
 
         TimeseriesFlattener(