!transition_variables
VariableName, VariableName, ...
...
!variables
VariableName, VariableName, ...
...
!transition_variables
VariableName, VariableName, ...
'Description of the variable...' VariableName
!transition_variables
VariableName, VariableName, ...
VariableName = Value
The !transition_variables
keyword starts a new declaration block for transition variables (i.e. endogenous variables); the names of the variables must be separated by commas, semi-colons, or line breaks. You can have as many declaration blocks as you wish in any order in your model file: They all get combined together when you read the model file in. Each variable must be declared (exactly once).
You can add descriptors to the variables (enclosed in single or double quotes, preceding the name of the variable); these will be stored in, and accessible from, the model object. You can also assign steady-state values to the variables straight in the model file (following an equal sign after the name of the variable); this is, though, rather rare and unnecessary practice because you can assign and change steady-state values more conveniently in the model object.
For each individual variable in a non-linear model, you can also decide if it is to be linearised or log-linearised by listing its name in the !log_variables
section.
!transition_variables
pie, 'Real output' Y
'Real exchange rate' Z = 1 + 1.05i;