I am using the dynare_sensitivity command to do the RMSE analysis for the posterior sample obtained by DYNARE’s Metropolis procedure (ppost = 1). When I choose morris = 2, I get the message
Error using save Variable 'rmse_mode' not found. Error in filt_mc_ (line 211) save([OutDir,filesep,fnamtmp,'.mat'], 'x', 'logpo2', 'likelihood', 'rmse_MC', 'rmse_mode','rmse_pmean', 'r2_MC', 'r2_mode','r2_pmean') Error in dynare_sensitivity (line 369) filt_mc_(OutputDirectoryName,options_gsa,dataset_);
When I choose morris = 1, I get the message but nothing else.
All parameter values in the specified ranges give unique saddle-path solution!
So I am wonder which value should the option morris be? if both are okay, what’s wrong when morris = 2? And when morris =1, as it seems nothing goes wrong, should it generate some figures as the output? Besides, is there any example for RMSE analysi with ppost = 1 to study? Attached is a simple NK model that demonstrate those messages and the data.
data.zip (1.1 KB)
nk_est.mod (1.37 KB)