Merge pull request #286 from flixOpt/v3.0.0/testing

Update Tests for years and scenarios

Author: FBumann

flixopt/elements.py

@@ -199,6 +199,7 @@ def __init__(
                 If the load-profile is just an upper limit, use relative_maximum instead.
             previous_flow_rate: previous flow rate of the flow. Used to determine if and how long the
                 flow is already on / off. If None, the flow is considered to be off for one timestep.
+                Currently does not support different values in different years or scenarios!
             meta_data: used to store more information about the Element. Is not used internally, but saved in the results. Only use python native types.
         """
         super().__init__(label, meta_data=meta_data)
@@ -305,7 +306,8 @@ def _plausibility_checks(self) -> None:
                 ]
             ):
                 raise TypeError(
-                    f'previous_flow_rate must be None, a scalar, a list of scalars or a 1D-numpy-array. Got {type(self.previous_flow_rate)}'
+                    f'previous_flow_rate must be None, a scalar, a list of scalars or a 1D-numpy-array. Got {type(self.previous_flow_rate)}.'
+                    f'Different values in different years or scenarios are not yetsupported.'
                 )
 
     @property

flixopt/features.py

@@ -171,21 +171,22 @@ def _do_modeling(self):
                 self.parameters.on_hours_total_max if self.parameters.on_hours_total_max is not None else np.inf,
             ),  # TODO: self._model.hours_per_step.sum('time').item() + self._get_previous_on_duration())
             short_name='on_hours_total',
-            coords=self.get_coords(['year', 'scenario']),
+            coords=['year', 'scenario'],
         )
 
         # 4. Switch tracking using existing pattern
         if self.parameters.use_switch_on:
             self.add_variables(binary=True, short_name='switch|on', coords=self.get_coords())
             self.add_variables(binary=True, short_name='switch|off', coords=self.get_coords())
 
-            ModelingPrimitives.state_transition_variables(
+            BoundingPatterns.state_transition_bounds(
                 self,
                 state_variable=self.on,
                 switch_on=self.switch_on,
                 switch_off=self.switch_off,
                 name=f'{self.label_of_model}|switch',
                 previous_state=self._previous_states.isel(time=-1) if self._previous_states is not None else 0,
+                coord='time',
             )
 
             if self.parameters.switch_on_total_max is not None:
@@ -408,7 +409,9 @@ def _do_modeling(self):
                 rhs = self.zero_point
             elif self._zero_point is True:
                 self.zero_point = self.add_variables(
-                    coords=self._model.get_coords(), binary=True, short_name='zero_point'
+                    coords=self._model.get_coords(('year', 'scenario') if self._as_time_series else None),
+                    binary=True,
+                    short_name='zero_point',
                 )
                 rhs = self.zero_point
             else:

flixopt/interface.py

@@ -246,8 +246,8 @@ def maximum_or_fixed_size(self) -> NonTemporalData:
 class OnOffParameters(Interface):
     def __init__(
         self,
-        effects_per_switch_on: Optional['NonTemporalEffectsUser'] = None,
-        effects_per_running_hour: Optional['NonTemporalEffectsUser'] = None,
+        effects_per_switch_on: Optional['TemporalEffectsUser'] = None,
+        effects_per_running_hour: Optional['TemporalEffectsUser'] = None,
         on_hours_total_min: Optional[int] = None,
         on_hours_total_max: Optional[int] = None,
         consecutive_on_hours_min: Optional[TemporalDataUser] = None,
@@ -339,15 +339,13 @@ def use_consecutive_off_hours(self) -> bool:
     @property
     def use_switch_on(self) -> bool:
         """Determines wether a Variable for SWITCH-ON is needed or not"""
-        return (
-            any(
-                param not in (None, {})
-                for param in [
-                    self.effects_per_switch_on,
-                    self.switch_on_total_max,
-                    self.on_hours_total_min,
-                    self.on_hours_total_max,
-                ]
-            )
-            or self.force_switch_on
+        if self.force_switch_on:
+            return True
+
+        return any(
+            param is not None and param != {}
+            for param in [
+                self.effects_per_switch_on,
+                self.switch_on_total_max,
+            ]
         )

flixopt/modeling.py

@@ -189,7 +189,7 @@ def expression_tracking_variable(
         name: str = None,
         short_name: str = None,
         bounds: Tuple[TemporalData, TemporalData] = None,
-        coords: List[str] = None,
+        coords: Optional[Union[str, List[str]]] = None,
     ) -> Tuple[linopy.Variable, linopy.Constraint]:
         """
         Creates variable that equals a given expression.
@@ -205,8 +205,6 @@ def expression_tracking_variable(
         if not isinstance(model, Submodel):
             raise ValueError('ModelingPrimitives.expression_tracking_variable() can only be used with a Submodel')
 
-        coords = coords or ['year', 'scenario']
-
         if not bounds:
             tracker = model.add_variables(name=name, coords=model.get_coords(coords), short_name=short_name)
         else:
@@ -223,86 +221,6 @@ def expression_tracking_variable(
 
         return tracker, tracking
 
-    @staticmethod
-    def state_transition_variables(
-        model: Submodel,
-        state_variable: linopy.Variable,
-        switch_on: linopy.Variable,
-        switch_off: linopy.Variable,
-        name: str,
-        previous_state=0,
-    ) -> Tuple[linopy.Constraint, linopy.Constraint, linopy.Constraint]:
-        """
-        Creates switch-on/off variables with state transition logic.
-
-        Mathematical formulation:
-            switch_on[t] - switch_off[t] = state[t] - state[t-1]  ∀t > 0
-            switch_on[0] - switch_off[0] = state[0] - previous_state
-            switch_on[t] + switch_off[t] ≤ 1  ∀t
-            switch_on[t], switch_off[t] ∈ {0, 1}
-
-        Returns:
-            variables: {'switch_on': binary_var, 'switch_off': binary_var}
-            constraints: {'transition': constraint, 'initial': constraint, 'mutex': constraint}
-        """
-        if not isinstance(model, Submodel):
-            raise ValueError('ModelingPrimitives.state_transition_variables() can only be used with a Submodel')
-
-        # State transition constraints for t > 0
-        transition = model.add_constraints(
-            switch_on.isel(time=slice(1, None)) - switch_off.isel(time=slice(1, None))
-            == state_variable.isel(time=slice(1, None)) - state_variable.isel(time=slice(None, -1)),
-            name=f'{name}|transition',
-        )
-
-        # Initial state transition for t = 0
-        initial = model.add_constraints(
-            switch_on.isel(time=0) - switch_off.isel(time=0) == state_variable.isel(time=0) - previous_state,
-            name=f'{name}|initial',
-        )
-
-        # At most one switch per timestep
-        mutex = model.add_constraints(switch_on + switch_off <= 1, name=f'{name}|mutex')
-
-        return transition, initial, mutex
-
-    @staticmethod
-    def sum_up_variable(
-        model: Submodel,
-        variable_to_count: linopy.Variable,
-        name: str = None,
-        bounds: Tuple[NonTemporalData, NonTemporalData] = None,
-        factor: TemporalData = 1,
-    ) -> Tuple[linopy.Variable, linopy.Constraint]:
-        """
-        SUms up a variable over time, applying a factor to the variable.
-
-        Args:
-            model: The optimization model instance
-            variable_to_count: The variable to be summed up
-            name: The name of the constraint
-            bounds: The bounds of the constraint
-            factor: The factor to be applied to the variable
-        """
-        if not isinstance(model, Submodel):
-            raise ValueError('ModelingPrimitives.sum_up_variable() can only be used with a Submodel')
-
-        if bounds is None:
-            bounds = (0, np.inf)
-        else:
-            bounds = (bounds[0] if bounds[0] is not None else 0, bounds[1] if bounds[1] is not None else np.inf)
-
-        count = model.add_variables(
-            lower=bounds[0],
-            upper=bounds[1],
-            coords=model.get_coords(['year', 'scenario']),
-            name=name,
-        )
-
-        count_constraint = model.add_constraints(count == (variable_to_count * factor).sum('time'), name=name)
-
-        return count, count_constraint
-
     @staticmethod
     def consecutive_duration_tracking(
         model: Submodel,
@@ -346,7 +264,7 @@ def consecutive_duration_tracking(
         duration = model.add_variables(
             lower=0,
             upper=maximum_duration if maximum_duration is not None else mega,
-            coords=model.get_coords(['time']),
+            coords=model.get_coords(),
             name=name,
             short_name=short_name,
         )
@@ -618,10 +536,55 @@ def scaled_bounds_with_state(
         )
         scaling_upper = model.add_constraints(variable <= scaling_variable * rel_upper, name=f'{name}|ub2')
 
-        big_m_upper = scaling_max * rel_upper
-        big_m_lower = np.maximum(CONFIG.modeling.EPSILON, scaling_min * rel_lower)
+        big_m_upper = rel_upper * scaling_max
+        big_m_lower = np.maximum(CONFIG.modeling.EPSILON, rel_lower * scaling_min)
 
         binary_upper = model.add_constraints(variable_state * big_m_upper >= variable, name=f'{name}|ub1')
         binary_lower = model.add_constraints(variable_state * big_m_lower <= variable, name=f'{name}|lb1')
 
         return [scaling_lower, scaling_upper, binary_lower, binary_upper]
+
+    @staticmethod
+    def state_transition_bounds(
+        model: Submodel,
+        state_variable: linopy.Variable,
+        switch_on: linopy.Variable,
+        switch_off: linopy.Variable,
+        name: str,
+        previous_state=0,
+        coord: str = 'time',
+    ) -> Tuple[linopy.Constraint, linopy.Constraint, linopy.Constraint]:
+        """
+        Creates switch-on/off variables with state transition logic.
+
+        Mathematical formulation:
+            switch_on[t] - switch_off[t] = state[t] - state[t-1]  ∀t > 0
+            switch_on[0] - switch_off[0] = state[0] - previous_state
+            switch_on[t] + switch_off[t] ≤ 1  ∀t
+            switch_on[t], switch_off[t] ∈ {0, 1}
+
+        Returns:
+            variables: {'switch_on': binary_var, 'switch_off': binary_var}
+            constraints: {'transition': constraint, 'initial': constraint, 'mutex': constraint}
+        """
+        if not isinstance(model, Submodel):
+            raise ValueError('ModelingPrimitives.state_transition_variables() can only be used with a Submodel')
+
+        # State transition constraints for t > 0
+        transition = model.add_constraints(
+            switch_on.isel({coord: slice(1, None)}) - switch_off.isel({coord: slice(1, None)})
+            == state_variable.isel({coord: slice(1, None)}) - state_variable.isel({coord: slice(None, -1)}),
+            name=f'{name}|transition',
+        )
+
+        # Initial state transition for t = 0
+        initial = model.add_constraints(
+            switch_on.isel({coord: 0}) - switch_off.isel({coord: 0})
+            == state_variable.isel({coord: 0}) - previous_state,
+            name=f'{name}|initial',
+        )
+
+        # At most one switch per timestep
+        mutex = model.add_constraints(switch_on + switch_off <= 1, name=f'{name}|mutex')
+
+        return transition, initial, mutex

flixopt/results.py

@@ -200,7 +200,7 @@ def __init__(
         self._flow_rates = None
         self._flow_hours = None
         self._sizes = None
-        self._effects_per_component = {'operation': None, 'invest': None, 'total': None}
+        self._effects_per_component = None
 
     def __getitem__(self, key: str) -> Union['ComponentResults', 'BusResults', 'EffectResults', 'FlowResults']:
         if key in self.components:
@@ -312,20 +312,24 @@ def filter_solution(
             startswith=startswith,
         )
 
-    def effects_per_component(self, mode: Literal['operation', 'invest', 'total'] = 'total') -> xr.Dataset:
-        """Returns a dataset containing effect totals for each components (including their flows).
-
-        Args:
-            mode: Which effects to contain. (operation, invest, total)
+    @property
+    def effects_per_component(self) -> xr.Dataset:
+        """Returns a dataset containing effect results for each mode, aggregated by Component
 
         Returns:
             An xarray Dataset with an additional component dimension and effects as variables.
         """
-        if mode not in ['operation', 'invest', 'total']:
-            raise ValueError(f'Invalid mode {mode}')
-        if self._effects_per_component[mode] is None:
-            self._effects_per_component[mode] = self._create_effects_dataset(mode)
-        return self._effects_per_component[mode]
+        if self._effects_per_component is None:
+            self._effects_per_component = xr.Dataset(
+                {
+                    mode: self._create_effects_dataset(mode).to_dataarray('effect', name=mode)
+                    for mode in ['operation', 'invest', 'total']
+                }
+            )
+            dim_order = ['time', 'year', 'scenario', 'component', 'effect']
+            self._effects_per_component = self._effects_per_component.transpose(*dim_order, missing_dims='ignore')
+
+        return self._effects_per_component
 
     def flow_rates(
         self,
@@ -580,7 +584,7 @@ def _compute_effect_total(
             total = xr.DataArray(np.nan)
         return total.rename(f'{element}->{effect}({mode})')
 
-    def _create_effects_dataset(self, mode: Literal['operation', 'invest', 'total'] = 'total') -> xr.Dataset:
+    def _create_effects_dataset(self, mode: Literal['operation', 'invest', 'total']) -> xr.Dataset:
         """Creates a dataset containing effect totals for all components (including their flows).
         The dataset does contain the direct as well as the indirect effects of each component.
 
@@ -590,24 +594,44 @@ def _create_effects_dataset(self, mode: Literal['operation', 'invest', 'total']
         Returns:
             An xarray Dataset with components as dimension and effects as variables.
         """
-        # Create an empty dataset
         ds = xr.Dataset()
+        all_arrays = {}
+        template = None  # Template is needed to determine the dimensions of the arrays. This handles the case of no shares for an effect
+
+        components_list = list(self.components)
 
-        # Add each effect as a variable to the dataset
+        # First pass: collect arrays and find template
         for effect in self.effects:
-            # Create a list of DataArrays, one for each component
-            component_arrays = [
-                self._compute_effect_total(element=component, effect=effect, mode=mode, include_flows=True).expand_dims(
-                    component=[component]
-                )  # Add component dimension to each array
-                for component in list(self.components)
-            ]
+            effect_arrays = []
+            for component in components_list:
+                da = self._compute_effect_total(element=component, effect=effect, mode=mode, include_flows=True)
+                effect_arrays.append(da)
+
+                if template is None and (da.dims or not da.isnull().all()):
+                    template = da
+
+            all_arrays[effect] = effect_arrays
+
+        # Ensure we have a template
+        if template is None:
+            raise ValueError(
+                f"No template with proper dimensions found for mode '{mode}'. "
+                f'All computed arrays are scalars, which indicates a data issue.'
+            )
+
+        # Second pass: process all effects (guaranteed to include all)
+        for effect in self.effects:
+            dataarrays = all_arrays[effect]
+            component_arrays = []
+
+            for component, arr in zip(components_list, dataarrays, strict=False):
+                # Expand scalar NaN arrays to match template dimensions
+                if not arr.dims and np.isnan(arr.item()):
+                    arr = xr.full_like(template, np.nan, dtype=float).rename(arr.name)
+
+                component_arrays.append(arr.expand_dims(component=[component]))
 
-            # Combine all components into one DataArray for this effect
-            if component_arrays:
-                effect_array = xr.concat(component_arrays, dim='component', coords='minimal')
-                # Add this effect as a variable to the dataset
-                ds[effect] = effect_array
+            ds[effect] = xr.concat(component_arrays, dim='component', coords='minimal')
 
         # For now include a test to ensure correctness
         suffix = {