Knut Rand

Attempt at a snakemake pipeline for benchmarking read-mappers: github.com/ivargr/mapping… Idea (as discussed by @krsahlin and others before) is that the community should work together on one clean pipeline for benchmarking instead of everyone making their own all the time.

@vsbuffalo Because of these points we decided in @BioNumPy to ditch pandas in favor of a lightweight bnpdataclass for biological data, which you define like python dataclasses, but allows numpy-like indexing of the rows:

One of my hotter takes is Python will never be a great data language until Pandas is replaced. This thread explains why really well: