Merge pull request #235 from dos-group/fix_#234

Fix #234

tests/test_signal.py

@@ -26,9 +26,7 @@ def hist_signal_single(self) -> vs.HistoricalSignal:
     @pytest.fixture
     def hist_signal_forecast(self) -> vs.HistoricalSignal:
         index = pd.date_range("2023-01-01T00:00:00", "2023-01-01T01:00:00", freq="20T")
-        actual = pd.DataFrame(
-            {"a": [1, 5, 3, 2], "b": [0, 1, 2, 3], "c": [4, 3, 2, 7]}, index=index
-        )
+        actual = pd.DataFrame({"a": [1, 5, 3, 2], "b": [0, 1, 2, 3]}, index=index)
 
         forecast_data = [
             ["2023-01-01T00:00:00", "2023-01-01T00:10:00", 2, 2.5],
@@ -294,12 +292,10 @@ def test_forecast_with_frequency(
         # Complicated because np.nan == np.nan is False
         assert forecast.keys() == expected.keys()
         assert all(
-            np.isnan(expected[k]) if np.isnan(forecast[k])
-            else forecast[k] == expected[k]
+            np.isnan(expected[k]) if np.isnan(forecast[k]) else forecast[k] == expected[k]
             for k in forecast.keys()
         )
 
-
     def test_forecast_fails_if_column_not_specified(self, hist_signal):
         with pytest.raises(ValueError):
             hist_signal.forecast(

vessim/signal.py

@@ -68,6 +68,13 @@ def __init__(
         fill_method: Literal["ffill", "bfill"] = "ffill",
         column: Optional[str] = None,
     ):
+        if isinstance(actual, pd.DataFrame) and forecast is not None:
+            if isinstance(forecast, pd.DataFrame):
+                if not actual.columns.equals(forecast.columns):
+                    raise ValueError("Column names in actual and forecast do not match.")
+            else:
+                raise ValueError("Forecast has to be a DataFrame if actual is a DataFrame.")
+
         self.default_column = column
         self._fill_method = fill_method
         # Unpack index of actual dataframe
@@ -356,7 +363,14 @@ def _resample_to_frequency(
         else:
             # Actual value is used for interpolation
             times = np.insert(times, 0, start_time)
-            data = np.insert(data, 0, self.now(start_time, column))
+            # https://github.com/dos-group/vessim/issues/234
+            # Use the length of the actual data to determine the column:
+            # self._actual is a dict[str, tuple[np.ndarray, np.ndarray]]
+            # -> every key is a column name
+            # -> if len(self._actual) == 1, _actual is based on pd.Series and column is None
+            data = np.insert(
+                data, 0, self.now(start_time, None if len(self._actual) == 1 else column)
+            )
             if resample_method == "ffill":
                 new_data = data[np.searchsorted(times, new_times, side="right") - 1]
             elif resample_method == "nearest":