[R-sig-ME] question about In mer_finalize(ans) : iteration limit reached without convergence (9)

Christopher Kurby kurbyc at gvsu.edu
Tue Jul 31 18:10:15 CEST 2012


Hello mixed modelers,

I have run some logistic mixed effect models recently and have received this error, "In mer_finalize(ans) : iteration limit reached without convergence (9)." My models are quite large. Here is one of them:

IndivChanges.Fine.slopes.lmer <- lmer(Bins ~ (Character + CharChar + CharObj + Space + Cause + Goal + Scene)*AgeCentered + TotalSpeed + (1+(Character + CharChar + CharObj + Space + Cause + Goal + Scene)|SubjNum) + (1+(Character + CharChar + CharObj + Space + Cause + Goal + Scene)|Clip), data=bins5000[bins5000$Grain == "Fine",], verbose=T, family=binomial)

No need to know the meaning of the regressors. As you can see, I have random effects terms that model random slopes associated with some of the fixed effects. I've read on related postings that the above error may be because the model is over-specified. This may be true because the model runs just fine if I model the intercepts only. Are the coefficients for the model with random slopes not to be trusted because of this warning? I have pasted the last few iterations below if that helps. 

Thanks much,
Chris

298:     32321.871: 0.731753 0.132115 0.137086  0.00000 0.00148956 0.0326407 0.168323 0.0525714 0.0972708 0.184065 0.0872438 0.133164 0.0684092 0.0539844 0.205175 0.851397 0.114391 0.381421 0.693689 0.158717  1.00432 0.866623 -0.206032 0.544031  1.19370 -1.50018 -0.190620 -0.0421279 -0.111096 -0.0849887 -0.174832 -0.318776 -0.346122 -0.371937 -1.33303 -0.742240 0.130250 0.116329  0.00000  0.00000 0.00425600  0.00000 0.147181 0.00293407 -0.461068 -0.557872 -0.716052 -1.19376 -1.26139  1.71680 -0.806458  1.86601 -0.594384 -0.250580 -0.0310011 -0.504158 -0.646994 0.574640 -0.586445 -0.207232 -0.950873 0.491170 0.216685 0.0412126 -0.159642 0.302386 -0.0298736 -0.242878 -0.398745 0.359931 -0.0315161 0.921738 -1.42477 0.317335 0.182896 0.376938 0.243562 0.377014 0.224896 -0.130288 0.00643905 0.0441177 0.000320275 -0.000801700 -0.00153798 0.000350855 0.000888126 -8.32440e-07 0.00235408
299:     32321.866: 0.731776 0.132129 0.137063  0.00000 0.00150241 0.0326668 0.168318 0.0525231 0.0972974 0.184061 0.0872274 0.133144 0.0684322 0.0539893 0.205210 0.851407 0.114443 0.381368 0.693660 0.158746  1.00431 0.866690 -0.206035 0.544054  1.19378 -1.50027 -0.190598 -0.0421015 -0.111112 -0.0849996 -0.174840 -0.318807 -0.346087 -0.372077 -1.33328 -0.742173 0.130232 0.116308  0.00000  0.00000 0.00423700  0.00000 0.147152 0.00295179 -0.461175 -0.557929 -0.716076 -1.19391 -1.26147  1.71705 -0.806294  1.86606 -0.594408 -0.250581 -0.0310267 -0.504231 -0.647073 0.574840 -0.586566 -0.207421 -0.951028 0.491237 0.216722 0.0412463 -0.159583 0.302432 -0.0300427 -0.242795 -0.398871 0.359955 -0.0315649 0.921779 -1.42473 0.317402 0.183039 0.376889 0.243582 0.377037 0.224942 -0.130254 0.00642826 0.0439955 0.000312418 -0.000800297 -0.00153592 0.000379481 0.000889780 8.91593e-06 0.00234924
300:     32321.482: 0.733417 0.133066 0.134693  0.00000 0.00437083 0.0390904 0.168895 0.0427666 0.101817 0.182775 0.0842960 0.129643 0.0729268 0.0538430 0.210895 0.853695 0.125644 0.370091 0.686940 0.165249  1.00100 0.881058 -0.206400 0.548838  1.21272 -1.52079 -0.185583 -0.0362476 -0.114788 -0.0873242 -0.176617 -0.325417 -0.338692 -0.403026 -1.38819 -0.727440 0.127952 0.111791  0.00000  0.00000 3.13331e-06  0.00000 0.141448 0.00714953 -0.485023 -0.570686 -0.721934 -1.22830 -1.28031  1.77333 -0.769850  1.87673 -0.599608 -0.251076 -0.0367640 -0.520249 -0.664602 0.619004 -0.613303 -0.249588 -0.985389 0.505882 0.224753 0.0486226 -0.146511 0.312606 -0.0677897 -0.224560 -0.426951 0.364995 -0.0425878 0.930731 -1.41525 0.334034 0.212488 0.367646 0.245462 0.382965 0.234438 -0.123606 0.00611488 0.0174095 0.000332695 -0.000623908 -0.00152730 0.000343676 0.000805522 4.24731e-05 0.00241407



More information about the R-sig-mixed-models mailing list