fdavidcl / s4r

Supervised complexity reduction autoencoder

s4r

Installation

For use

In an R session:

remotes::install_github("fdavidcl/s4r")

For development

In a shell:

$ git clone https://github.com/fdavidcl/s4r.git
$ cd s4r
$ R

In the R session:

remotes::install_github("fdavidcl/dcme")
remotes::install_bioc("graph")
remotes::install_bioc("Rgraphviz")
devtools::install_dev_deps()

Dependencies

• Tensorflow (up to 2.2; 2.4 is not supported)

s4r's Issues

Revise LSSVM loss

• Is max(., 0) necessary? . should be in [0,2] in general.
• Divide svm_term by two? Seems appropriate in the LS formulation