uwmisl / poretitioner

https://misl.cs.washington.edu
Other
3 stars 1 forks source link

Code profiling: NanoporeTER vs poretitioner #94

Closed kdoroschak closed 3 years ago

kdoroschak commented 3 years ago

For the paper & thesis, comparing NanoporeTER publication methods with poretitioner methods in terms of disk size, memory usage, and time.

Currently profiling for old NanoporeTER methods code while we plumb through the pipeline for the docker image.

https://docs.google.com/spreadsheets/d/1J1YeUN1hPUyFLpvtwd0F4dBVebyBFcY0-CF7dRdxgoE/edit#gid=0

Notable: I think the batch size was set very small for NanoporeTER code so the lab folks could spin up a bunch of these at once when leaving for the day, rather than having to queue them, so it takes a long time.