update sorting and schema explanations

core/src/validation.jl

@@ -102,6 +102,7 @@ sort_by_id_state_level(row) = (row.node_id, row.control_state, row.level)
 sort_by_priority(row) = (row.node_id, row.priority)
 sort_by_priority_time(row) = (row.node_id, row.priority, row.time)
 sort_by_subgrid_level(row) = (row.subgrid_id, row.basin_level)
+sort_by_id_name(row) = (row.node_id, row.name)
 
 # get the right sort by function given the Schema, with sort_by_id as the default
 sort_by_function(table::StructVector{<:Legolas.AbstractRecord}) = sort_by_id
@@ -110,6 +111,7 @@ sort_by_function(table::StructVector{BasinProfileV1}) = sort_by_id_level
 sort_by_function(table::StructVector{UserDemandStaticV1}) = sort_by_priority
 sort_by_function(table::StructVector{UserDemandTimeV1}) = sort_by_priority_time
 sort_by_function(table::StructVector{BasinSubgridV1}) = sort_by_subgrid_level
+sort_by_function(table::StructVector{DiscreteControlCompoundvariableV1}) = sort_by_id_name
 
 const TimeSchemas = Union{
     BasinTimeV1,

docs/core/usage.qmd

@@ -628,10 +628,18 @@ DiscreteControl is implemented based on [VectorContinuousCallback](https://docs.
 
 ## DiscreteControl / compoundvariable
 
+The compound variable schema defines linear combinations of variables which can be used in conditions. This means that
+this schema defines new variables with the given name that look like
+$$
+\text{weight}_1 * \text{variable}_1 + \text{weight}_2 * \text{variable}_2 + \ldots,
+$$
+
+which can be for instance an average or a difference of variables. If a variable comes from a time-series, a look ahead $\Delta t$ can be supplied.
+
 column            | type     | unit    | restriction
 ----------------- | -------- | ------- | -----------
 node_id           | Int32    | -       | sorted
-name              | String   | -       | -
+name              | String   | -       | sorted per node_id
 listen_node_type  | String   | -       | known node type
 listen_node_id    | Int32    | -       | sorted per node_id
 variable          | String   | -       | must be "level" or "flow_rate", sorted per listen_node_id
@@ -640,7 +648,8 @@ look_ahead        | Float64  | $s$     | Only on transient boundary conditions,
 
 ## DiscreteControl / condition
 
-The condition schema defines conditions of the form 'the discrete_control node with this node id listens to whether the given variable of the node with the given listen feature id is grater than the given value'. If the condition variable comes from a time-series, a look ahead $\Delta t$ can be supplied.
+The condition schema defines conditions of the form 'the discrete_control node with this node id listens to whether the given variable of the node with the given listen node id is greater than the given value'. If the condition variable comes from a time-series, a look ahead $\Delta t$ can be supplied.
+Compound variables from the schema above can also be used, in which case the listen node ID and look ahead in this schema are ignored.
 
 column            | type     | unit    | restriction
 ----------------- | -------- | ------- | -----------

python/ribasim/ribasim/config.py

@@ -290,13 +290,11 @@ class ManningResistance(MultiNodeModel):
 class DiscreteControl(MultiNodeModel):
     compoundvariable: TableModel[DiscreteControlCompoundvariableSchema] = Field(
         default_factory=TableModel[DiscreteControlCompoundvariableSchema],
-        json_schema_extra={"sort_keys": ["name"]},
+        json_schema_extra={"sort_keys": ["node_id", "name"]},
     )
     condition: TableModel[DiscreteControlConditionSchema] = Field(
         default_factory=TableModel[DiscreteControlConditionSchema],
-        json_schema_extra={
-            "sort_keys": ["node_id", "listen_node_id", "variable", "greater_than"]
-        },
+        json_schema_extra={"sort_keys": ["node_id"]},
     )
     logic: TableModel[DiscreteControlLogicSchema] = Field(
         default_factory=TableModel[DiscreteControlLogicSchema],