I'm running siglasso with some samples that may be problematic, and I get errors like
Error: from glmnet Fortran code (error code 10000); All penalty factors are <= 0
I'm wondering if siglasso should just skip these (and fill the relevant sample data with NA). At the same time, is there any way to find why these samples are problematic? A few of these had too few variants (and were promptly removed), but others have reasonable numbers (50+).
I'm running siglasso with some samples that may be problematic, and I get errors like
I'm wondering if siglasso should just skip these (and fill the relevant sample data with
NA). At the same time, is there any way to find why these samples are problematic? A few of these had too few variants (and were promptly removed), but others have reasonable numbers (50+).