Page 1 of 1

mode_compute=6, still positive definite matrix problem

PostPosted: Thu Apr 28, 2011 5:24 pm
by anne
Hey all,

I had the usual error message telling me that the Hessian matrix was not semi definite positive, so I use the "mode_compute=6" command. But I still get the error message :

??? Error using ==> chol
Matrix must be positive definite.

Error in ==> gmhmaxlik at 232
dd = transpose(chol(CovJump));

Error in ==> dynare_estimation_1 at 552
[xparam1,PostVar,Scale,PostMean] = ...

Error in ==> dynare_estimation at 62
dynare_estimation_1(var_list,varargin{:});

Error in ==> argentina at 240
dynare_estimation(var_list_);

Error in ==> dynare at 132
evalin('base',fname) ;

What can I do ??

Thanks for the help !

Anne.