Closed GFabien closed 6 months ago
Attention: 39 lines
in your changes are missing coverage. Please review.
Comparison is base (
17347b7
) 94.39% compared to head (55c51f1
) 95.02%.:exclamation: Current head 55c51f1 differs from pull request most recent head b0cb631. Consider uploading reports for the commit b0cb631 to get more accurate results
Files | Patch % | Lines |
---|---|---|
R/rgcca_inner_loop.R | 72.54% | 14 Missing :warning: |
R/rgcca_outer_loop.R | 81.57% | 14 Missing :warning: |
R/block_postprocess.R | 47.05% | 9 Missing :warning: |
R/block_init.R | 85.71% | 2 Missing :warning: |
:exclamation: Your organization needs to install the Codecov GitHub app to enable full functionality.
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
Use S3 classes to handle the different types of blocks. We refer to primal blocks for blocks having more rows than columns and dual blocks in the opposite case. So far, the following types of blocks are handled:
tau = sparsity = 1
(block),tau = sparsity = 1
(dual_block),tau != 1
(primal_regularized_block),tau != 1
(dual_regularized_block),sparsity != 1
(sparse_block). These different classes of blocks allow the implementations of RGCCA and SGCCA. The formerrgccad
andsgcca
functions have been merged into thergcca_outer_loop
function, and the formerrgccak
andsgccak
functions have been merged into thergcca_inner_loop
function. This last function creates the block objects with their corresponding classes and calls the associated methods to initialize, update, and post-process the blocks.In addition, it is now possible to specify a custom connection matrix when there is a response block.