I am trying to run a mixed model logistic regression with participants
nested within state with a certain amount of covariates. Here is what my
model looks like:
m1 <- lmer(Overweight ~ age + factor(A_RACE_G) + Prevalance +
HH_Income_Dicot + Unemployment_Rate + Intensity_effect + (1 | state2) +
GSD_EFFECT + FMA_EFFECT + BMI_EFFECT + DBS_EFFECT + NSM_EFFECT + NCF_EFFECT
+ ACF_EFFECT + NES_EFFECT + AMS_EFFECT + FTS_EFFECT + PRI_EFFECT +
PUI_EFFECT + MLR_EFFECT + PES_EFFECT + PAS_EFFECT + PFA_EFFECT + HES_EFFECT
+ SRS_EFFECT + SWP_EFFECT + TAX_EFFECT + WBP_EFFECT, data = reduced, family
= binomial (link = "logit"), REML = FALSE)
My main concern is that the model does not run when I add in ",
weights=A_FINALWT" which is the survey weighting variable.
1: In eval(expr, envir, enclos) :
non-integer #successes in a binomial glm!
2: glm.fit: algorithm did not converge
3: glm.fit: fitted probabilities numerically 0 or 1 occurred
That is the error I receive.
The model will run if I use the continuous BMI as the dependent variable and
run it without family = binomial. However, we want a logistic regression,
not a linear one.
What is it that is not working with weighting the model by my sampling
weight?
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
