Stephan Daus

I find that mutate also drops attributes whenever there is a a function call inside mutate, whereas simle mathematics works. ``` r library(dplyr) #> #> Attaching package: 'dplyr' #> The...

Any reason why no suggesting this for the dplyr package? I think many people have similar issues when joining, and I prefer the idea of having all join functions in...

I strongly support recognizing {labelled} in tabyls. I am about to introduce tidyverse, janitor and more to many R-newbies who face a lot of survey data, but this limitation makes...

My bug seems related, though is not about exporting but simply importing a Stata-dataset. Was quite surprised to see that `haven::read_dta()` successfully read value labels, whereas `rio::import()` does not.

Is this of any help? https://www.statmodel.com/download/JSM2007000746.pdf

Yeah, I recognized it was on the model-based side rather than design-based, but hoped you could magically extract what you needed... These ones below are more on design-based inference, but...

Ah, that seems indeed like a sufficient solution for my colleagues. Thanks!

A bit on the side but a warning to passers-by: Sharepoint/OneDrive still has a limit of 400 characters on its paths, so when working on these folders, sync might stop...

If I can chip in on your first statement, if I understand you correctly. I think it would lead to a nightmare if each package can implement their own formatting...

That's nice, though I think you deserve more. If they provided a month's salary you could perhaps attract more assistance from the other contributors? (I know you are busy yourself...