pappewaio
commented
2 years ago Test case data:
RSID B SE Z P N AFREQ
rs4819391_G 0.233253438056324 0.127172429896691 1.83415098890387 0.0668444836995174 1397 NA
rs11089128_G -0.057230689617411 0.0707237246183849 -0.809214869921227 0.418529557835942 1397 NA
rs7288972_C -0.0603896529952508 0.0561292812495112 -1.07590283807129 0.282156749586622 NA 0.166428060128848
rs2032141_A 2.04506156636745 0.987859595989405 2.07019456476422 NA 1397 0.000357909806728704
rs11912265_C 0.0730128523576906 0.148272187537675 NA 0.622496796832969 1397 0.0164638511095204
rs4321465_T 0.0184879350623236 NA 0.0558985901144536 0.955430598683162 1397 0.00322118826055834
rs11167319_G NA 0.0539111323459842 -0.946776039745882 0.343916880859673 1397 0.134216177523264
rs8138488_C -0.0636612849960759 0.0377416410861792 -1.68676515286423 0.0918721012352699 1397 0.374015748031496
rs2186521_A 0.0549469615715985 0.060835465298906 0.903206070696177 0.366572470028154 1397 0.10773085182534
Test case execution:
cat tmp_test_data.txt | ./build/r-stats-c-streamer --functionfile functiontestfile.txt --skiplines 1 --index 1 --pvalue 5 --beta 2 --standarderror 3 --Nindividuals 6 --zscore 4 --allelefreq 7 --statmodel lin | head | column -t
When one value is NA, it is better to return NA instead if inf, which now is the case for inference of N, using an incomplete list of allele frequencies, which happens when adding 1000 genomes AF:s