gEconpy.model.build.statespace_from_gcn

gEconpy.model.build.statespace_from_gcn(gcn_path, simplify_blocks=True, simplify_tryreduce=True, simplify_constants=True, infer_steady_state=True, verbose=True, on_unused_parameters='raise', log_linearize=True, not_loglin_variables=None, show_errors=True, filter_type='standard')

Build a symbolic DSGE state-space model from a GCN file.

Unlike model_from_gcn, this returns a DSGEStateSpace whose steady-state and linearized system are represented as pytensor graphs parameterized by the model’s free parameters. This is the entry point for Bayesian estimation via PyMC.

Parameters:

gcn_pathstr or Path

Path to the GCN file.

simplify_blocksbool, default True

Simplify block equations during parsing.

simplify_tryreducebool, default True

Eliminate user-marked tryreduce variables.

simplify_constantsbool, default True

Fold constant “variables” into equations.

infer_steady_statebool, default True

Propagate analytical steady-state solutions through identities.

verbosebool, default True

Print a build report on completion.

on_unused_parametersstr, default 'raise'

How to handle unused parameters: 'raise', 'warn', or 'ignore'.

log_linearizebool, default True

Whether to log-linearize the model.

not_loglin_variableslist of str, optional

Variable names to exclude from log-linearization.

show_errorsbool, default True

Pretty-print parse errors to stderr.

filter_typestr, default 'standard'

Kalman-filter variant to use in the underlying PyMCStateSpace.

Returns:

DSGEStateSpace

A symbolic state-space model ready for estimation.