Geweke always doesn't converge

[GoogleCodeExporter]

if GLM doesn't converge, would it be possible to recheck again after one more 
sample? would one sample (or even 10) make a difference to the convergence of 
the GLM? 

Original issue reported on code.google.com by Sibon...@gmail.com on 17 Nov 2012 at 12:38

[GoogleCodeExporter]

well, at least if we check we can get "some" result sooner. i.e. if you check 
for geweke convergence every sample instead of once every 1000 samples you're 
likely to be able to tell if it has converged straight away.

ideally, this is what should be done:

chain length = 10000 check - GLM not converged. nextcheckchainlength=10001, 
progress=NaN
chain length = 10001 check - GLM not converged. nextcheckchainlength=10002, 
progress=NaN
chain length = 10000 check - GLM converged. 
nextcheckchainlength=20000(proportional scale to progress) , progress=50%
chain length = 20000 check - GLM not converged. 
nextcheckchainlength=20001(proportional scale to progress) , progress=50%

Original comment by Sibon...@gmail.com on 20 Nov 2012 at 12:36

[GoogleCodeExporter]

Original comment by Sibon...@gmail.com on 27 Nov 2012 at 2:11

• Changed state: Started

[GoogleCodeExporter]

Original comment by steven...@gmail.com on 27 Nov 2012 at 5:36

[GoogleCodeExporter]

Original comment by Sibon...@gmail.com on 28 Nov 2012 at 1:07

• Changed state: Fixed