1 test fails on PowerPC: group_counts(c(1:200), c(0.55)) not equal to c(110, 90) about groupdata2

[barracuda156]

Yes, everything passes now:

--->  Testing R-groupdata2
* using log directory ‘/opt/local/var/macports/build/_opt_PPCSnowLeopardPorts_R_R-groupdata2/R-groupdata2/work/groupdata2-ca20bcb17f053fa0408d82999a3a3ee054773c76/groupdata2.Rcheck’
* using R version 4.3.0 (2023-04-21)
* using platform: powerpc-apple-darwin10.0.0d2 (32-bit)
* R was compiled by
    gcc-mp-12 (MacPorts gcc12 12.2.0_2) 12.2.0
    GNU Fortran (MacPorts gcc12 12.2.0_2) 12.2.0
* running under: OS X Snow Leopard 10.6
* using session charset: UTF-8
* checking for file ‘groupdata2/DESCRIPTION’ ... OK
* this is package ‘groupdata2’ version ‘2.0.2.9000’
* package encoding: UTF-8
* checking package namespace information ... OK
* checking package dependencies ... OK
* checking if this is a source package ... OK
* checking if there is a namespace ... OK
* checking for executable files ... OK
* checking for hidden files and directories ... OK
* checking for portable file names ... OK
* checking for sufficient/correct file permissions ... OK
* checking whether package ‘groupdata2’ can be installed ... OK
* checking installed package size ... OK
* checking package directory ... OK
* checking DESCRIPTION meta-information ... OK
* checking top-level files ... OK
* checking for left-over files ... OK
* checking index information ... OK
* checking package subdirectories ... OK
* checking R files for non-ASCII characters ... OK
* checking R files for syntax errors ... OK
* checking whether the package can be loaded ... OK
* checking whether the package can be loaded with stated dependencies ... OK
* checking whether the package can be unloaded cleanly ... OK
* checking whether the namespace can be loaded with stated dependencies ... OK
* checking whether the namespace can be unloaded cleanly ... OK
* checking loading without being on the library search path ... OK
* checking dependencies in R code ... OK
* checking S3 generic/method consistency ... OK
* checking replacement functions ... OK
* checking foreign function calls ... OK
* checking R code for possible problems ... OK
* checking Rd files ... OK
* checking Rd metadata ... OK
* checking Rd cross-references ... OK
* checking for missing documentation entries ... OK
* checking for code/documentation mismatches ... OK
* checking Rd \usage sections ... OK
* checking Rd contents ... OK
* checking for unstated dependencies in examples ... OK
* checking files in ‘vignettes’ ... WARNING
Files in the 'vignettes' directory but no files in 'inst/doc':
  ‘automatic_groups_with_groupdata2.Rmd’
  ‘cross-validation_with_groupdata2.Rmd’
  ‘description_of_groupdata2.Rmd’ ‘introduction_to_groupdata2.Rmd’
  ‘styles.css’ ‘time_series_with_groupdata2.Rmd’
* checking examples ... OK
* checking for unstated dependencies in ‘tests’ ... OK
* checking tests ...
  Running ‘testthat.R’
 OK
* checking for unstated dependencies in vignettes ... OK
* checking package vignettes in ‘inst/doc’ ... WARNING
Directory 'inst/doc' does not exist.
Package vignettes without corresponding single PDF/HTML:
  ‘automatic_groups_with_groupdata2.Rmd’
  ‘cross-validation_with_groupdata2.Rmd’
  ‘description_of_groupdata2.Rmd’
  ‘introduction_to_groupdata2.Rmd’
  ‘time_series_with_groupdata2.Rmd’
* checking running R code from vignettes ...
  ‘automatic_groups_with_groupdata2.Rmd’ using ‘UTF-8’... OK
  ‘cross-validation_with_groupdata2.Rmd’ using ‘UTF-8’... OK
  ‘description_of_groupdata2.Rmd’ using ‘UTF-8’... OK
  ‘introduction_to_groupdata2.Rmd’ using ‘UTF-8’... OK
  ‘time_series_with_groupdata2.Rmd’ using ‘UTF-8’... OK

[LudvigOlsen]

Perfect! :-)

[LudvigOlsen]

PowerPC! Wow!

I guess powerpc rounds 0.55 differently?

Not sure 1) how to check that (have no powerpc), 2) what to do if it is the case, 3) whether it's actually problematic - besides reproducibility of analyses. But if rounding is different, that should cause other problems and perhaps powerpc isn't the place to run reproducibility checks? :-)

Let me know your thoughts!

[barracuda156]

@LudvigOlsen Thank you for responding!

Since there is no Fortran code, I believe, rounding is done via libm functions? Rounding mode can be controlled, but default one should be “towards nearest”.

If there is a specific test case which I could use to diagnose the issue, I can do that. (And yeah, rounding is tricky and time to time leads to test errors. Been there, eh.)

[LudvigOlsen]

Hi @barracuda156

I just looked at the code used in this test.
It is basically saying (0.55 / 1.0) * 200 == 110. The output (110) is floor rounded. The only way I case see this become 109 is if 1.0 is >1.0 (e.g. 1.0000000...0000001 or something).

The 1.0 in this instance is calculated as 1 - (0 / 200) (all integers). Which should give us 1.0. Perhaps you can check if that gives the expected result on PowerPC?

I've constrained the 1.0 variable to be maximum of 1.0, which could fix this issue. You can test it by installing the latest version of the master branch (`devtools::install_github("ludvigolsen/groupdata2").

[LudvigOlsen]

@barracuda156

Once you have tested that this solution (hopefully) works, I will make a new release :-)

[barracuda156]

@LudvigOlsen Sorry for a delay, I will test this tonight.