Mode_compute = 9 is not working


#1

Hello people,

Initially I was using mode_compute 6, but that was not identifiying the mode properly (checked through mode plots) and MCMC were not converging. So I started using mode_compute 9.

But now I am getting, warnings:

  1. Initial sigma much smaller than the given bounded interval.
  2. Non-finite fitness range

And then finally execution stops with the error:
error using chol
MAtrix must be positive definite

Any suggestions in which direction I should proceed to resolve this?

w_19_BE_with_SC_AR1.mod (10.6 KB)
final_file.xlsx (13.1 KB)


#2

I tried mode_compute=5 and got

The following parameters are at the prior bound: SE_e_sc, SE_e_omss, SE_e_gi_me

You should try to understand why your data pushes these standard errors to the bound.