Hi everybody,

I want to open a Stata Dataset in R and used readstata13 to keep the labels. This is not possible in a similar way with haven. Now there is the problem that concert.factors=TRUE keeps the labels but converts the numbers starting from 1. That means my negative missings from the Stata Dataset and also codings with 0 are changed starting from 1. If the numbers e.g. where -897, -898, 1, 2, 3, they are in R transformed to 1,2,3,4,5. So the missings are suddenly on numbers 1 and 2 instead of the negative ones.

This is my code:

```
pp2014p_stata13 <- read.dta13("phd2014_p_d_4-0-0.dta",
convert.factors=TRUE,
encoding = NULL,
nonint.factors = TRUE,
generate.factors=TRUE,
missing.type=TRUE,
convert.dates=FALSE,
select.rows=NULL,
)
```

I now read in a publication by Martin Elff, that this problem can somehow be solved by using the memisc package. But I have no idea how this should work.

I would try this line for the missings

valid.values (pp2014p_stata13>0)

but I´m not sure how this could be integrated in the code above. If I run this directly after readstata13, the changes have been already done. Is there a possibility to include code for both packages in one?

Thank you for useful hints and tips in advance!