These are chat archives for dereneaton/ipyrad

12th
Jan 2016
Deren Eaton
@dereneaton
Jan 12 2016 00:00
Internal indels, that is.
Isaac Overcast
@isaacovercast
Jan 12 2016 00:19
I like max_indels_locus, and i like what you did with all the step7 filtering params, make them all share a common format.
Also, i recruited our first alpha tester today. Glenn Seeholzer from LSU, he's a good dude, got a pile of single end GBS to analyse. i'm going to put him on my dev box and let him hack, good way to find bugs.
Isaac Overcast
@isaacovercast
Jan 12 2016 00:25
I want to add a 'preview' option to main. I want to support the ability of users to do ipyrad -p params.txt --preview and have it just run through everything really quick to make sure everything is kosher. Behind the scenes it'll just do step1(preview) and push the clipped data down the pipe.
Deren Eaton
@dereneaton
Jan 12 2016 00:34
Great idea. And good to hear about getting Glen on board. I haven't tested any SE GBS yet, so I expect he'll run into problems really quick. We should definitely do testing on0 pairddrad data, that seems really popular, and its the kind that crashes most often in pyrad.
Isaac Overcast
@isaacovercast
Jan 12 2016 20:08
Lol. i fixed preview mode for step1. Originally i was calling preview and subsetting data after parallel_chunker, so on real data it still took all night. I moved preview to subset the incoming raw data prior to parallel_chunker. Now it takes ~30 seconds. Much more "preview-like".
oh yeah, hackersonly dict is nice. I had to update the length of the default truncated preview file, can just update the assembly rather than recompile? Very choice.
Deren Eaton
@dereneaton
Jan 12 2016 21:30
That makes sense. Nice!
Deren Eaton
@dereneaton
Jan 12 2016 21:41
Copying this here as a reminder about new vsearch and lower case letters: torognes/vsearch#147