irf.msbsvar {MSBVAR} | R Documentation |
Simulates a posterior of impulse response functions (IRF) by
Monte Carlo integration. This can handle MSBSVAR models estimated
with the msbsvar
function. Simulations of IRFs
from the Bayesian model utilize the posterior estimates for that model.
irf.msbsvar(gibbs, msbsvar, nsteps)
gibbs |
Output from the gibbs.msbsvar function. This is
the posterior for the MSBSVAR model
|
msbsvar |
Fitted msbsvar posterior mode from a call to
msbsvar .
|
nsteps |
Integer, number of steps or periods in the impulse response horizon. |
This is a beta function. It averages over the posterior for the
regimes and generates a set of regime-specific impulse responses.
There is one set of responses produced for each endogenous variable in
each regime. The identification of the A(0) for the system
is set up in the msbsvar
function. See that for details.
MSBSVAR:
An irf.MSBSVAR class object object that is the array of
impulse response samples for the Monte Carlo samples
impulse |
A list of h draws X nsteps X (m*m) arrays of the impulse responses. The list
if ordered by regime. These are IRFs using the associated B-SVAR
model in msbsvar and the posterior
A(0). |
Eventually this function will be wrapped into the classing mechanism for
estimation and plotting for IRFs. This is described in the
mc.irf
documentation.
Patrick T. Brandt
gibbs.msbsvar
, msbsvar
,
irf
, mc.irf