[R-pkg-devel] noLD error in CRAN checks in package 'HACSim'

[Uwe Ligges]

On 15.05.2019 19:50, Jarrett Phillips wrote:
> Hello,
> 
> My package has recently been posted to CRAN, however, there is an
> additional issue of noLD on tests on x86_64 Linux with R-devel.
> 
> I haven't yet received an email from CRAN maintainers on resolving the
> issue.
> 
> However, I would like to submit an updated package version with improved
> documentation and thought that I could fix this error as well.
> 
> Any ideas on how to solve the below problem?
> 
> I suspect it has something to do with numerical precision of estimates.

Right, apparently there is some test that tries to confirm that "probs 
must sum to 1" and that is not exactly true on platforms without long 
doubles. You can test for that via all.equal(), perhaps you have to 
adjust the tolerance.

Best,
Uwe Ligges



> 
> -----
> 
> * checking examples ... ERROR
> Running examples in ‘HACSim-Ex.R’ failed
> The error most likely occurred in:
> 
>> ### Name: HAC.simrep
>> ### Title: Run a simulation of haplotype accumulation curves for
>> ###   hypothetical or real species
>> ### Aliases: HAC.simrep
>>
>> ### ** Examples
>>
>>
>>    ## Simulate hypothetical species ##
>>
>>    N <- 100 # total number of sampled individuals
>>    Hstar <- 10 # total number of haplotypes
>>    probs <- rep(1/Hstar, Hstar) # equal haplotype frequency distribution
>>
>>    HACSObj <- HACHypothetical(N = N, Hstar = Hstar , probs = probs,
> +   filename = "output") # outputs a CSV file called "output.csv"
>>
>>    ## Simulate hypothetical species - subsampling ##
>>    HACSObj <- HACHypothetical(N = N, Hstar = Hstar, probs = probs,
> +   perms = 1000, p = 0.95, subsample = TRUE, prop = 0.25,
> +   conf.level = 0.95, filename = "output")
>>
>>    ## Simulate hypothetical species and all paramaters changed - subsampling ##
>>    HACSObj <- HACHypothetical(N = N, Hstar = Hstar, probs = probs,
> +   perms = 10000, p = 0.90, subsample = TRUE, prop = 0.15, conf.level = 0.95,
> +   filename = "output")
>>
>>    HAC.simrep(HACSObj) # runs a simulation
> 
>   Simulating haplotype accumulation...
> 
> 
>    |
>    |
>    |   0%Error in HAC.sim(N = envr$N, Hstar = envr$Hstar, probs =
> envr$probs, perms = envr$perms,  :
>    probs must sum to 1
> Calls: HAC.simrep -> HAC.sim
> Execution halted
> * checking PDF version of manual ... OK
> * DONE
> 
> Status: 1 ERROR
> See
>    ‘/data/gannet/ripley/R/packages/tests-noLD/HACSim.Rcheck/00check.log’
> for details.
> 
> 
> Thanks.
> 
> 
> Cheers,
> 
> Jarrett
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-package-devel using r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-package-devel
>