Closed tretherington closed 4 years ago
This has been resolved by reworking the code primarily to:
This issue is being closed as the speed up is sufficient for current purposes. If further speed gains are required then parallelisation could be a good option, but we decided this was probably best done by the user level by applying the function to subsets of the desired test points.
The find-simplex.R function isn't executing fast enough to be useful in application - though I can confirm that it is working correctly in 2 and 3 dimensions.
I suspect the nested for loops initiating on lines 62 and 66 are the issue, so it would be great to explore how this process could be optimised through using things like the apply function in R to avoid the for loops and multi-core processing.
It would be ideal if we could keep as much of the code in R so that it is readily accessible, but we could also consider moving the whole process into a C script.