Bug in as.mids2 about mice HOT 7 CLOSED

Hi thanks. The as.mids2() function was as a temporary fixup created long ago. I have added several fixups, and was therefore wondering whether you get the same behaviour of the as.mids() function in a recent version of mice.
Stef.

from mice.

We have found the same issue using either function. I posted the code using as.mids2 because I had understood it was an "updated" fix, but even with as.mids the same issue is happening.
Thanks for your response.

from mice.

I am not able to run your example since it seems to require SPSS. At some point, there is a hard-coded file C:/FILEPATH.sav, which is don't have. I cannot see what the nature of the problem is.

Is your finding perhaps related to exporting to and importing from SPSS, rather than being an as.mids() problem? There is a function in mice called mids2spss() that is meant to export mids objects to SPSS.

from mice.

I apologize that the example cannot be ran without creating the SPSS file from the .sps document. The FILEPATH location should be pointing to the SPSS file (which should be your working directory).

The following syntax should skip the SPSS file creation altogether.
Thank you.
example2.txt

from mice.

That's curious. In mice 2.36 running your code gives me identical counts. Which version of mice produces your problem, and on what platform?
reproduce.pdf

from mice.

We have been running version 2.30 on windows 10. When installing/updating the package there is no indication of version 2.36, so perhaps this might have been addressed in the updated version and we are just not caught up for some reason.

from mice.

OK thanks. mice 2.30 is the latest on CRAN. It appears that current version of as.mids() works as expected. I am going to close, but please let me know if problem persist. Stef.

from mice.