[cudapoa] unifying closely related numerous parameters related to POA graph length

NVIDIA-Genomics-Research / GenomeWorks

SDK for GPU accelerated genome assembly and analysis

https://clara-parabricks.github.io/GenomeWorks/

Apache License 2.0

286 stars 76 forks source link

[cudapoa] unifying closely related numerous parameters related to POA graph length #521

Closed r-mafi closed 4 years ago

r-mafi commented 4 years ago

Currently there are different variables that are closely related to graph length, such as max_nodes_per_window, max_matrix_graph_dimension, and their banded versions. It makes life much easier to deal with only one parameter instead.

r-mafi commented 4 years ago

banded and non-banded have been unified into one variable. There has been also changes that distinct max_nodes_per_window (now changed to max_nodes_per_graph) and max_matrix_graph_dimension (now changed to matrix_graph_dimension). The latter is specifically used for score-matrix buffer size, while the former is used for other POA arrays buffer size. These changes are addressed in PR#519