R：eval中的错误（predvars，data，env）：数字'envir'arg在multgee包的intrinsic.pars（）中长度不为1

Question

I have a data frame df with the following data:

family_id_cass   twin_order  ls75cat
  10026          1           2
  10026          2           2
  10071          1           3
  10071          2           3
  10132          1           1
  10132          2           2

I would like to run intrinsic.pars() , so what I am doing is the following:

intrinsic.pars(df$ls75cat, df$family_id_cass, df$twin_order, rscale="ordinal") but i get the following error:

Error in eval(predvars, data, env) : numeric 'envir' arg not of length one

What am I doing wrong?

Answer 1

intrinsic.pars() seems to be written with a very temperamental interface which effectively ignores most of the parameters. (I suspect this was meant to be an internal function.) In any case, you have to provide a data.frame to the argument 'data', which includes variables y , id , and time :

df <- read.csv(text=
"family_id_cass,twin_order,ls75cat
10026,1,2
10026,2,2
10071,1,3
10071,2,3
10132,1,1
10132,2,2")

newData<-data.frame(y=df$ls75cat,
                    id=df$twin_order,
                    time=df$family_id_cass)

# either of these calls will do:
intrinsic.pars(y,newData,id,time,rscale="ordinal")
intrinsic.pars(data=newData,rscale="ordinal")

#> [1] -3.498327e-07 -1.143481e-10 -5.162827e+00
#> Warning message:
#> glm.fit: fitted rates numerically 0 occurred