heibl / ips

R Interfaces to Phylogenetic Software
11 stars 3 forks source link

Update mafft.R #3

Closed Andreas-Bio closed 7 years ago

Andreas-Bio commented 7 years ago

I had permament crashes due to low memory (I am editing .fasta >500 MB). I switched these lines and it worked again.

heibl commented 7 years ago

Thanks for your input. Can you say, which of the functions (scan or read.fas) caused the crash? Commit is merged, anyway, thanks again!

Andreas-Bio commented 7 years ago

Thanks! It was not a crash really. Maybe I should have described that in more detail. I am not a data scientist, sorry. My RAM just kept filling up slowly each time I called the mafft function to the point where it became so slow I had to force quit R because my mouse was hardly moving any more. I had once a similar problem with read.fas (I was using a loop to read 100 fasta files and my RAM also filled up), that's why I changed it to read.dna. However ,I figured changing scan might also help. I did not specifically test both changes individually, all I can say is that the RAM consumption dropped by 40% after the loop, compared to the version before.