Estimating and Smoothing/Decomposing Shocks with different datasets

Estimation ML/Bayesian estimation
1

Hi all,
I am trying to estimate a model on pre-COVID data and then using the resulting estimates to obtain series for smoothed shocks and historical shock decompositions for the post-COVID period.
I know that I could do this using two .mod files, but this seems cumbersome and I was wondering if there is a more practical way of doing it in the same mod-file.
So far, I have been unsuccessful. Here is what I tried: say that ‘data_precovid.mat’ includes data from 1984q1 to 2019q4, and ‘data_full.mat’ includes data from 1984q1 to 2023q4. I tried to set up the following commands in my mod file:

estimation(datafile=data_precovid, …);
calib_smoother(datafile=data_full,…);
shock_decomposition(datafile=data_full,…);

This runs fine, but the resulting oo_.SmoothedVariables and oo_.shock_decomposition only have series with 144 observations, meaning that they were done with ‘data_precovid.mat’ only. Is there anything that I am missing, or is this just not possible in a single .mod file?

Thanks in advance,

2

Without seeing the files, I can only speculate. But it may be necessary to explicitly set nobs.

3

Thanks a lot Johannes. Setting ‘nobs’ explicitly worked!