broadinstitute / catch / 294

95%

master: 94%

push

travis-ci-com

Parallelize creation of distance matrix for clustering

Filling in the condensed distance matrix, in the cluster module, is (by
far) the bottleneck of clustering input sequences. But this can be
easily parallelized. This commit uses a multiprocessing.Pool to break up
entries in the condensed matrix, and has separate processes fill in
those entries.

1516 of 1709 branches covered (88.71%)

58 of 58 new or added lines in 2 files covered. (100.0%)

5193 of 5459 relevant lines covered (95.13%)

2.85 hits per line