topic Re: Data steps , in SAS Programming

Data steps ,

epimph69 — Thu, 20 Feb 2014 10:11:09 GMT

Hello everybody ,

1) I have this message in my log , after usinf proc mi if anyone may tell me about it

"WARNING: The EM algorithm (MLE) fails to converge after 200 iterations. You can increase the

number of iterations (MAXITER= option) or increase the value of the convergence

criterion (CONVERGE= option).

WARNING: The EM algorithm (posterior mode) fails to converge after 200 iterations. You can

increase the number of iterations (MAXITER= option) or increase the value of the

convergence criterion (CONVERGE= option)."

2 ) I have 500 variables in my datasets and I wish to work with 10variables only , and I specified the 10 ten variables and I have the same results

Thank advance for your replies

Re: Data steps ,

Amir — Thu, 20 Feb 2014 13:55:26 GMT

Hi,

This warning has been mentioned in the forums a few times, but no answer seems to have been reached:

https://communities.sas.com/community/support-communities/search.jspa?peopleEnabled=true&userID=&containerType=&container=&q=fails+to+converge+after+200+iterations

If you don't get any responses then it might be worth contacting SAS Technical Support.

Regards,

Amir.

Re: Data steps ,

Cynthia_sas — Thu, 20 Feb 2014 14:13:35 GMT

...And, searching Tech Support for relevant notes seems to indicate that the message is procedure dependent. I found notes about PROC NLIN and PROC GENMOD (paper).

cynthia

22969 - Why does my nonlinear model fail to converge? How can I get it to converge?

http://www2.sas.com/proceedings/sugi28/270-28.pdf

Re: Data steps ,

stat_sas — Thu, 20 Feb 2014 14:24:04 GMT

I think some of the variables being used for this analysis are highly correlated. Please check correlation among 10 variables and retain only uncorrelated variables for this analysis. I hope this will solve the problem.

Regards,