bwgs.cv - output issue

[IvanaOsCro]

First of all, I have to say that I find the BWGS package for R really useful and user friendly. Especially for people like me that are new to genomic selection.

But, I'm facing a problem when using one of the functions and I hope there is a solution to it.

Namely, when using the „bwgs.cv“ function together with „pop.reduct.method“ set to „OPTI“ (CDmean optimisation) I'm not getting mean CV correlation for any of nTimes.

So, the code runs properly, CV correlation for each of the nFolds is calculated, but for each of nTimes CV correlation I'm getting NA. And then of course, at the end of the process mean correlation and standard deviation CVs are NA.

This is a part of the output I'm getting:

cv correlation for fold 48 is: 0.585

cv correlation for fold 49 is: 0.592

cv correlation for fold 50 is: 0.711

                          |****************************************| 100%

Mean CV correlation for time 1 and 50 folds is: NA

This happens only when I use CDmean optimization. When I don't use optimization at all or when I use random optimization, this problem doesn't occur. Also, it occurs whether I use my data set or INRA47 example dataset.

Do you know why this is happening and, perhaps, how to avoid it and get mean correlation and standard deviation CVs in the final output?