aggregate input for generalized linear e.g. logit
Posted: Sun Aug 09, 2020 10:06 am
congratulations to developer - great package
it would be really helpful if input for binomial distributions could include and aggreegatee format for proportions aswell as the 0,1 format, as in R glmer
Examplein R: model model <- glmer(cbind(f0, f1) ~ a1_23 * b1 +(1|pno), data=ldf, family=binomial(link="logit"))
is equivalent to model <- glmer(bo1~ a1_23 * b1 +(1|pno), data=ldf, family=binomial(link="logit"))
where b01 is a binary variable with value 0 or 1
f0 is aggregate number of 0 values
f1 is aggregate number of 1 values
Have data set witeh proportions for different maxima rangin from 4 to 200 -expansion for b01 makes a very long file
it would be really helpful if input for binomial distributions could include and aggreegatee format for proportions aswell as the 0,1 format, as in R glmer
Examplein R: model model <- glmer(cbind(f0, f1) ~ a1_23 * b1 +(1|pno), data=ldf, family=binomial(link="logit"))
is equivalent to model <- glmer(bo1~ a1_23 * b1 +(1|pno), data=ldf, family=binomial(link="logit"))
where b01 is a binary variable with value 0 or 1
f0 is aggregate number of 0 values
f1 is aggregate number of 1 values
Have data set witeh proportions for different maxima rangin from 4 to 200 -expansion for b01 makes a very long file