Merge c1b24dd into fe21a28

flexmeasures/data/models/planning/storage.py

@@ -214,9 +214,8 @@ def _prepare(self, skip_validation: bool = False) -> tuple:  # noqa: C901
                 query_window=(start, end),
                 resolution=resolution,
                 beliefs_before=belief_time,
-                default_value_attribute="production_capacity",
-                default_value=convert_units(power_capacity_in_mw, "MW", sensor.unit),
-                method="upper",
+                fallback_attribute="production_capacity",
+                max_value=convert_units(power_capacity_in_mw, "MW", sensor.unit),
             )
         if sensor.get_attribute("is_strictly_non_negative"):
             device_constraints[0]["derivative max"] = 0
@@ -230,9 +229,8 @@ def _prepare(self, skip_validation: bool = False) -> tuple:  # noqa: C901
                 query_window=(start, end),
                 resolution=resolution,
                 beliefs_before=belief_time,
-                default_value_attribute="consumption_capacity",
-                default_value=convert_units(power_capacity_in_mw, "MW", sensor.unit),
-                method="upper",
+                fallback_attribute="consumption_capacity",
+                max_value=convert_units(power_capacity_in_mw, "MW", sensor.unit),
             )
 
         # Apply round-trip efficiency evenly to charging and discharging

flexmeasures/data/models/planning/tests/test_solver.py

@@ -1269,8 +1269,9 @@ def set_if_not_none(dictionary, key, value):
             False,
             None,
             None,
-            # from the flex model field 'production-capacity' (a sensor)
-            [-0.2] * 4 * 4 + [-0.3] * 4 * 4 + [-8] * 16 * 4,
+            # from the flex model field 'production-capacity' (a sensor),
+            # and when absent, defaulting to the max value from the power sensor attribute capacity_in_mw
+            [-0.2] * 4 * 4 + [-0.3] * 4 * 4 + [-10] * 16 * 4,
             # from the power sensor attribute 'consumption_capacity'
             [0.5] * 24 * 4,
         ),
@@ -1282,8 +1283,9 @@ def set_if_not_none(dictionary, key, value):
             None,
             # from the power sensor attribute 'consumption_capacity'
             [-8] * 24 * 4,
-            # from the flex model field 'consumption-capacity' (a sensor)
-            [0.25] * 4 * 4 + [0.15] * 4 * 4 + [0.5] * 16 * 4,
+            # from the flex model field 'consumption-capacity' (a sensor),
+            # and when absent, defaulting to the max value from the power sensor attribute capacity_in_mw
+            [0.25] * 4 * 4 + [0.15] * 4 * 4 + [10] * 16 * 4,
         ),
         (
             "Test battery with dynamic power capacity",
@@ -1305,7 +1307,7 @@ def set_if_not_none(dictionary, key, value):
             # from the flex model field 'production-capacity' (a quantity)
             [-1] * 24 * 4,
             # from the power sensor attribute 'consumption_capacity' (a quantity)
-            [0.5] * 24 * 4,
+            [2] * 24 * 4,
         ),
         (
             "Test battery",

flexmeasures/data/models/planning/utils.py

@@ -373,41 +373,32 @@ def get_continuous_series_sensor_or_quantity(
     unit: ur.Quantity | str,
     query_window: tuple[datetime, datetime],
     resolution: timedelta,
-    default_value_attribute: str | None = None,
-    default_value: float | int | None = np.nan,
+    fallback_attribute: str | None = None,
+    max_value: float | int | None = np.nan,
     beliefs_before: datetime | None = None,
-    method: str = "replace",
 ) -> pd.Series:
     """
     Retrieves a continuous time series data from a sensor or quantity within a specified window, filling
-    the missing values from an attribute (`default_value_attribute`) or default value (`default_value`).
-
-    Methods to fill-in missing data:
-        - 'replace' missing values are filled with the default value.
-        - 'upper' clips missing values to the upper bound of the default value.
-        - 'lower' clips missing values to the lower bound of the default value.
+    the missing values from a given `fallback_attribute` and making sure no values exceed `max_value`.
 
     :param quantity_or_sensor:      The quantity or sensor containing the data.
     :param actuator:                The actuator from which relevant defaults are retrieved.
     :param unit:                    The desired unit of the data.
     :param query_window:            The time window (start, end) to query the data.
     :param resolution:              The resolution or time interval for the data.
-    :param default_value_attribute: Attribute for a default value if data is missing.
-    :param default_value:           Default value if no attribute or data found.
+    :param fallback_attribute:      Attribute serving as a fallback default in case no quantity or sensor is given.
+    :param max_value:               Maximum value (also replacing NaN values).
     :param beliefs_before:          Timestamp for prior beliefs or knowledge.
-    :param method:                  Method for handling missing data: 'replace', 'upper', 'lower', 'max', or 'min'.
     :returns:                       time series data with missing values handled based on the chosen method.
-    :raises: NotImplementedError:   If an unsupported method is provided.
     """
 
     _default_value = np.nan
 
-    if default_value_attribute is not None:
+    if fallback_attribute is not None:
         _default_value = get_quantity_from_attribute(
             entity=actuator,
-            attribute=default_value_attribute,
+            attribute=fallback_attribute,
             unit=unit,
-            default=default_value,
         )
 
     time_series = get_series_from_quantity_or_sensor(
@@ -418,15 +409,16 @@ def get_continuous_series_sensor_or_quantity(
         beliefs_before=beliefs_before,
     )
 
-    if method == "replace":
+    # Use default as fallback
+    if quantity_or_sensor is None:
         time_series = time_series.fillna(_default_value)
-    elif method == "upper":
-        time_series = time_series.fillna(_default_value).clip(upper=_default_value)
-    elif method == "lower":
-        time_series = time_series.fillna(_default_value).clip(lower=_default_value)
-    else:
-        raise NotImplementedError(
-            "Method `{method}` not supported. Please, try one of the following: `replace`, `max`, `min` "
-        )
+
+    # Apply upper limit
+    time_series = nanmin_of_series_and_value(time_series, max_value)
 
     return time_series
+
+
+def nanmin_of_series_and_value(s: pd.Series, value: float) -> pd.Series:
+    """Perform a nanmin between a Series and a float."""
+    return s.fillna(value).clip(upper=value)