ggpairs_custom: Plot pairwise correlations of multiple features

In brightchan/cjbmisc:

Description

Plot a big complex correlation map of every features in a dataframe or specific features (feat.plot). Usage: ggpairs_custom(df,plot.it=TRUE) (all features). Or p<-ggpairs_custom(df,c("Ploidy","Purity","EGFR","Smoker"))
then ggsave("plot.pdf",p)

Usage

ggpairs_custom(
  ggdf,
  feat.plot = NULL,
  colist = comp_hm_colist_full,
  sig.col = c("white", "thistle1", "orchid1"),
  plot.it = FALSE,
  signif.cutoff = 0.05,
  return.pvalue = FALSE,
  verbose = FALSE,
  ...
)

Arguments

ggdf	dataframe with rows of samples and columns of features
feat.plot	character vector of the features to be plotted. If NULL then all columns will be plotted
colist	a named list of named vector of colors to be used for each feature. Missing ones are automatically handled.
sig.col	vector of 3 colors for nonsignif, lower and higher bound of the significance.
plot.it	if FALSE, no plot printed, only return the object. Use ggsave(plot=p$plot) to plot.
signif.cutoff	cutoff point for significant p-value. A colored (sig.col[3]) frame will be shown to significant plots.
return.pvalue	whether to return a pvalue matrix. If yes, it has to plot it to generate the values (to a new device)
verbose	whether to show the plotting pairs for debugging
...	pass to ggpairs

Value

If return.pvalue=FALSE:"gg" "ggmatrix" plotable and ggsavable object; if TRUE: a list of two: $plot is the plot as when return.pvalue=FALSE; $p.value is the table of correlation pvalues

brightchan/cjbmisc documentation built on Nov. 5, 2021, 4:12 p.m.