This work is main branch version for one of optimization works done in nextflow branch.
It reduces the running time of 5 minutes for demo.vcf.gz on elpaso server, this running time may differ by your machine.
For more details, please refer the link to https://github.com/hyunhwan-bcm/AI_MARRVEL/pull/9
Applied scalar multiplication to vector, not on matrix
Used indexing technique
Transformed from float64 to float32
Changed Output
There's subtle change on outputs, but I think they are ignorable.
>>> pd.testing.assert_series_equal(a["diffuse_Phrank_STRING"], b["diffuse_Phrank_STRING"])
Series values are different (0.04513 %)
This work is main branch version for one of optimization works done in nextflow branch. It reduces the running time of 5 minutes for demo.vcf.gz on elpaso server, this running time may differ by your machine. For more details, please refer the link to https://github.com/hyunhwan-bcm/AI_MARRVEL/pull/9
Changed Output
There's subtle change on outputs, but I think they are ignorable.