Page 1 of 1

about the too narrow posterior distribution

PostPosted: Mon Apr 15, 2013 3:31 am
by jinyou_yin
Dear all,
I applied bayes estimation with command:
estimation(order=1, datafile=level_fs2000_simul, nobs=200, mode_compute=6, mh_replic=2000, mh_nblocks=2, mh_jscale=0.8);

but I got a very narrow posterior distribution. I want to fixed it by change the option of mh_jscale=0.8, But when I change the option of mh_jscale, it says: Some element of Newton direction isn't finite. Jacobian maybe singular or there is a problem with initial values ? how can i fix this problem?
Thanks for answers,

Best,

Jinyou