npyLoad() not working with the last version of R

[HomereAMK]

Hello Dirk,

I think the npyLoad function is not working properly for the last version of R (R version 4.2.1 (2022-06-23) -- "Funny-Looking Kid")

 rm(list = ls(all = TRUE))
  library(tidyverse)
  library(RcppCNPy)

genome_selection <- npyLoad("trial_inv_pcangsd_sca4_e2.selection.npy")

I can't send the input as the format is not supported by github...

[eddelbuettel]

Let's slow down for a second. First, there is no need for the universally disliked rm(list=ls(all-TRUE)) is posted code. Second, we do not need library(tidyverse). That leaves with loading the package and a command I cannot reproduce. There is nothing I can work with here.

On the other hand if you look at https://cran.r-project.org/web/checks/check_results_RcppCNPy.html you see that the package still tests perfectly fine with the current version of R.

I suspect you have some 'non-standard' data in that npy file. You need to figure out what it is, and if you can, create a 'mock version' I can work with. We had prior reports here before about data fields not representable in R, maybe see old issues at https://github.com/eddelbuettel/rcppcnpy/issues?q=is%3Aissue+is%3Aclosed.

[HomereAMK]

Thanks for your super quick reply

Well understood, the reason why I'm so adamant about it is that the same exact code (yes, with universally disliked rm(list=ls(all-TRUE)) and the beautiful library(tidyverse) is working fine and give the output that I attached to this message for an older version of R. Homere.txt

[eddelbuettel]

:man_shrugging:

Still not reproducible, still no change in the package, still passes at CRAN. I can't help you here, and I tried to help you help yourself.

[HomereAMK]

Lovely, I'll help myself with my help and close the issue 👍🏾