RC_plsda_loading: RC_plsda_loading
In bvittrant/RCommon: Basic functions

Description Usage Arguments Details Value Author(s) References Examples

View source: R/RC_plsda_loading.R

Perform a PLS-DA in the same way as RC_PLSDA1 but return the loadings plot of the desired componant.

1	RC_plsda_loading(Data, Design, ...)

`Data`	Your initial dataframe containing the gene expression or microarray value for examples.
`Design`	Your dataframe descrining your sample. It should match the Data dataframe.
`Name_project`	A title name for your project. You should define it as an env variable at the beginning of your work.
`comp`	The componant you want to plot (1,2,3 ...).
`choice1`	The main feature you want to use in the PLS-DA. It should be an existing column from your Design dataframe.
`choice2`	A feature you want to use to do a multilevel PLS-DA. This will be the feature used to correct the first.

If only choice1 is provided then do a single level PLS-DA. If choice2 is provided then do a multi-level PLS-DA corrected for choice2.

Return a graph object

Benjamin Vittrant

http://mixomics.org/

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

## The function is currently defined as
RC_PLSDA1 = function(Data, Design, Name_Project, choice1, choice2 = NULL, ...){
  # Choice1 & choice2, any column from your design file
  # If only choice1 is provided then it's a single level PLS-DA
  # If choice 2 is provided it's a pls-da on choice1 corrected for choice2
  tmp1 = choice1
  tmp2 = choice2

  # PLSDA
  if(is.null(tmp2)) {
    # Simple level PLSDA mixOmics
    plsda = plsda(Data, Design[,tmp1], ncomp = 10, ...)
    p = plotIndiv(plsda , comp = 1:2,
                  group = Design[,tmp1], ind.names = FALSE,
                  ellipse = TRUE, legend = TRUE, title = paste("Simple PLSDA - ",tmp1, sep=""))
    return(p)
  }

  if(!is.null(tmp2)){
    # Multilevel PLSDA mixOmics
    plsda = plsda(Data, Design[,tmp1], ncomp = 10, multilevel = Design[,tmp2], ...)
    p = plotIndiv(plsda , comp = 1:2,
                  group = Design[,tmp1], ind.names = FALSE,
                  ellipse = TRUE, legend = TRUE,
                  title = paste("Multilevel PLSDA - ",tmp1," vs ", tmp2, sep="")
    )
    return(p)
  }
}