n samples in sample_group and datafile are different error with CRAN install, not reproduced after installation from github

[joshuamschmidt]

Get an error that n samples in sample_group and datafile are different when they are not:

unload('smartsnp')
remove.packages('smartsnp')   
pacman::p_load('smartsnp','data.table','ggplot2','ggrepel',install = T, update = F)
group_id <- fam$phase
length(group_id)

[1] 9411

SP <- which(fam$phase != "kg")
gfile <- 'hidden.traw'
sm.pca <- smart_pca(snp_data = gfile, sample_group = group_id, missing_value = 9, missing_impute = "mean",
                    scaling = "drift", program_svd = "RSpectra", pc_axes = 2, 
                    sample_project = SP, pc_project = 1:2)

Imported 126351 SNP by 9411 sample genotype matrix
Time elapsed: 0h 0m 44s
Filtering data...
Error in smart_pca(snp_data = gfile, sample_group = group_id, missing_value = 9, :
length(sample_group) should be equal to number of samples in dataset: computation aborted

This is an uninformative error, as it doesn't report the expected number of samples given by sample_group variable. I added code to do this (see separate pull request #5.), but upon installing this branch, the error went away. This is reproducible with the main branch:

unload('smartsnp')
remove.packages('smartsnp')
devtools::install_github("ChristianHuber/smartsnp")
library('smartsnp')
sm.pca <- smart_pca(snp_data = gfile, sample_group = group_id, missing_value = 9, missing_impute = "mean",
                    scaling = "drift", program_svd = "RSpectra", pc_axes = 2, 
                    sample_project = SP, pc_project = 1:2)

Imported 126350 SNP by 9411 sample genotype matrix
Time elapsed: 0h 0m 8s
Filtering data...
126350 SNPs included in PCA computation
2504 samples included in PCA computation
6907 samples projected after PCA computation
Completed data filtering
Time elapsed: 0h 0m 14s
Scanning for invariant SNPs...
Scan complete: no invariant SNPs found
Time elapsed: 0h 0m 28s
Checking for missing values...
Scan completed: no missing values found
Time elapsed: 0h 0m 29s
Scaling values by SNP...
Centering and scaling by drift dispersion...
Completed scaling using drift
Time elapsed: 0h 0m 30s
Computing singular value decomposition using RSpectra...
Completed singular value decomposition using RSpectra
Time elapsed: 0h 0m 51s
Extracting eigenvalues and eigenvectors...
Eigenvalues and eigenvectors extracted
Time elapsed: 0h 0m 51s
Projecting ancient samples onto PCA space
PCA space = PC1PC2
Completed ancient sample projection
6907 ancient samples projected
Time elapsed: 0h 2m 5s
Tabulating PCA outputs...
Completed tabulations of PCA outputs...

I note that the number of SNPs differs by 1. Has the function to read in traw files changed?

session_info()

setting value
version R version 4.2.1 (2022-06-23 ucrt)
os Windows 10 x64 (build 18363)
system x86_64, mingw32
ui RStudio
language (EN)
collate English_Australia.utf8
ctype English_Australia.utf8
tz Australia/Adelaide
date 2022-08-01
rstudio 2022.02.3+492 Prairie Trillium (desktop)
pandoc NA

[ChristianHuber]

Thanks for the fix of the error message!
No, we didn't change the function for reading traw files. Not sure why there is one SNP less.