Page 1 of 1

Bootstrapping problem

Posted: Tue Aug 16, 2011 3:16 am
by Florian
Hello,

I am trying to estimate a VAR with GLS in EVIEWs and am performing bootstrapping in order to get a confidence interval for the variables of interest. Every time I run the model I get the following error message:

Positive or non-negative argument to function expected in computation of group weight (variance) in equation equation gls_{%gr}_y2.ls(wgt=cxdiag,cx=f,cov=cxwhite) gdp_z gdp_z(-1) gdp_z(-2) gdp_z(-3) gdp_z(-4) gc_z(-1) gc_z(-2) gc_z(-3) gc_z(-4) c

where gc_z is the new bootstrapped data.

I am attaching the program I wrote, it would be great if somebody hat time to take a brief look at it.

Thank you

Florian

Re: Bootstrapping problem

Posted: Tue Aug 16, 2011 3:19 am
by Florian
Attached the program!

Re: Bootstrapping problem

Posted: Tue Aug 16, 2011 7:51 am
by EViews Gareth
probably need the data too.