Closed abuts closed 7 months ago
as by-side job of maintaining Horace-3 (Re #1587) it worth enabling sorting pixels using mex code. Profiling suggests it is 20 times faster then Matlab code.
The PR enables mex code for sorting pixels and adds separate (and more) unit tests for it.
In addition to that the PR contains minor fix to ClusterWrapper, (one row, commit https://github.com/pace-neutrons/Horace/pull/1589/commits/32fbe4e7b6ba8c31202694a56917f2992b5a7a9d) which fixes the issue identified while running parpool cluster on IDAAAS
Solves #1018
as by-side job of maintaining Horace-3 (Re #1587) it worth enabling sorting pixels using mex code. Profiling suggests it is 20 times faster then Matlab code.
The PR enables mex code for sorting pixels and adds separate (and more) unit tests for it.
In addition to that the PR contains minor fix to ClusterWrapper, (one row, commit https://github.com/pace-neutrons/Horace/pull/1589/commits/32fbe4e7b6ba8c31202694a56917f2992b5a7a9d) which fixes the issue identified while running parpool cluster on IDAAAS
Solves #1018