jschoeley commented 6 years ago

I implemented Q <- t(W %*% B) in RcppEigen and achieved a doubling of speed in pclm2D.

Resources for what I did:

see https://stackoverflow.com/questions/35923787/fast-large-matrix-multiplication-in-r for the idea of using RcppEigen to speed up matrix operations
see http://adv-r.had.co.nz/Rcpp.html (Using Rcpp in a package) for the required manual change of the NAMESPACE file
see https://github.com/hadley/devtools/issues/1675 for a strange bug in testthat which lets tests fail when using devtools::test() although they work perfectly when using devtools::test_dir()

codecov-io commented 6 years ago

Codecov Report

Merging #1 into master will increase coverage by 0.06%. The diff coverage is 100%.

@@            Coverage Diff             @@
##           master       #1      +/-   ##
==========================================
+ Coverage   93.33%   93.39%   +0.06%     
==========================================
  Files           8        9       +1     
  Lines         525      530       +5     
==========================================
+ Hits          490      495       +5     
  Misses         35       35

Impacted Files	Coverage Δ
src/tMatMult.cpp	`100% <100%> (ø)`
R/pclm_fit.R	`98.88% <100%> (ø)`	:arrow_up:

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 9eeb836...3f540b4. Read the comment docs.

mpascariu commented 6 years ago

Hi Jonas,

Please open a new pull request to the devel branch. Avoid sending your .Rproj file since is not necessary for building the package from GitHub.

I would also be very happy if you could provide a c++ function for the entire loop in pclm_fit.R (R29-44). If you do not have time submit as it is and I will look into it.

Thanks a lot for your contribution. Let's get this ready for CRAN. Marius

mpascariu / ungroup

Speed Up pclm2D with RcppEigen #1

Codecov Report