Sumtable Vars redux
qvrb opened this issue · comments
when i try to call sumtable im getting the error Error in round(results[y], digits = max(digits[y], 1)) :
non-numeric argument to mathematical function
the failed call st(data,vars = c('responses', 'offers', 'bestoffer','meanoffer','name','polite','price','black'))
if i call st(data,vars = c('responses', 'offers', 'bestoffer','meanoffer','name','polite','price')) it works fine
so that implies that black is non-numeric except
str(data$black) returns TRUE
dbl+lbl [1:1200] 0, 0, 1, 0, 1, 1, 0, 0, 0, 1, 1, 1, 0, 1, 0, 1, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1,...
@ label : chr "Black"
@ format.stata: chr "%9.0g"
@ labels : Named num [1:2] 0 1
..- attr(*, "names")= chr [1:2] "Other" "Black"
head(data)
head(data)
ad responses offers bestoffer meanoffer name polite price texttype black tattoo white
1: 55 0 0 NA NA NA NA 130 0 0 0 1
2: 59 1 1 95 95.0 1.0 1.0 110 0 0 0 1
3: 60 2 0 NA NA 1.0 1.0 90 0 1 0 0
4: 65 0 0 NA NA NA NA 110 0 0 0 1
5: 66 5 2 90 82.5 0.2 0.6 90 0 1 0 0
6: 67 2 1 75 75.0 0.0 0.0 110 0 1 0 0
Hard for me to help much without having the data on hand. But it looks like the issue is the use of value labels, which the package does try to support if they come from the haven package but also can cause shenanigans in corner cases. In your particular case I'd imagine you could fix the problem by converting black to a proper factor variable, I think sjlabelled::as_factor() does this.
Hi,
I've attached the sta file but renamed it as a csv as github won't allow sta uploads
Thanks a million for your prompt response
Regards
library(haven library(vtable) library(sjlabelled) data <- read_dta("c:\\temp\\data_doleacstein.dta") st(data,vars = c('responses', 'offers', 'bestoffer','meanoffer','name','polite','price','offers',sjlabelled::as_factor('black'))))
which gives me
which is not quite what i was expecting since i get a 1 for the data in the black label
That CSV file is corrupted and won't work; it looks like you just changed the file extension to CSV instead of using export delimited or something. In any case, if you did convert to CSV I suspect the problem would go away. The problem here is definitely the use of haven and it importing value labels from a Stata file, which wouldn't happen with a CSV.
Your recent post doesn't work because vars only accepts the names of columns, you can't do calculations in it. If you first create hte factor variable, using perhaps mutate(factor_black = sjlabelled::as_factor('black')) or something like that (if you're using tidyverse), and then using 'factor_black' in your vars argument, I suspect that would work.