Skip to content

Latest commit

 

History

History
35 lines (23 loc) · 2.25 KB

README.md

File metadata and controls

35 lines (23 loc) · 2.25 KB
Raw

FDR-corrected SCCA R scripts

The R scripts in the examples directory perform the simulation studies described in Gossmann et. al. FDR-Corrected Sparse Canonical Correlation Analysis with Applications to Imaging Genomics (2017). Code to generate visualizations from the simulation results, as presented in the paper, is provided in this repository as well.

The R scripts have been written with the intention of running them on the high performance computing cluster Cypress at Tulane University, which uses the Slurm resource management system.

R package

Many functions (see the R directory), which are used to implement the simulations in the examples directory, are provided in form of an R package. The R package FDRcorrectedSCCA can be installed using devtools in R (see below). However, this should not be necessary in order to run the scripts from the examples directory, because within those scripts all functions from the FDRcorrectedSCCA package are loaded with devtools::load_all().

However, if you have trouble running the example scripts (ahem Windows ahem...), try installing the required functions as an R package following the instructions below. In that case the line devtools::load_all() has to be replaced with library(FDRcorrectedSCCA) in the example codes.

The R package FDRcorrectedSCCA can be installed form a running R session with:

# install.packages("devtools") # if devtools is not installed
library(devtools)
install_github("agisga/FDRcorrectedSCCA")

Development workflow - Manual installation of the R package

  1. Install the devtools R package (if you don't have it installed already):

    install.packages("devtools")

  2. Open the file FDRcorrectedSCCA.Rproj in RStudio (if no Rproj file is available, you can create the Rproj file from RStudio via File > New Project > Existing Directory).

  3. Then run

    devtools::document()

  4. Still in RStudio press CTRL-Shift-B. (If any of steps 1-4 fail, you are probably missing some other R packages, see the Error or Warning messages. Install the missing packages and try again from step 2.)