plot_feat_subtype: Plot multiple features of subtypes

In brightchan/cjbmisc:

Description

Plot percentage barplot or violin/boxplot for discrete or continuous variables of each subtype. Return a list of ggplots that can be arranged into figure panel with ggarrange.
Usage: plist <- plot_feat_subtype(df,"Stage",c("Ploidy","Purity","EGFR","Smoker"))
then p <- ggpubr::ggarrange(plotlist=plist,align="hv")
ggsave("file.pdf",p)

Usage

plot_feat_subtype(
  plotdf,
  subtype,
  feat.plot,
  cont.plot = "boxplot",
  cont.test = "wilcox.test",
  disc.test = "both",
  p.label = NULL,
  p.symnum.args = list(),
  cont.show.mean = FALSE,
  show.count = TRUE,
  cont.col = comp_hm_colist_full$disc,
  disc.col = comp_hm_colist_full,
  disc.width = 0.8,
  cont.width = 0.9,
  disc.border = TRUE,
  cont.border = TRUE,
  anno.textsize = 4,
  cont.anno.vjust = 0.2,
  disc.anno.vjust = 0.04,
  aspect.ratio = 1.4,
  legend.position = "bottom",
  highlight.signif = TRUE,
  highlight.signif.col = "red",
  signif.cutoff = 0.05,
  plot.signif.only = F,
  ...
)

Arguments

plotdf	dataframe with rows of samples and columns of features, one column should contain the subtype info.
subtype	column name of the subtype
feat.plot	character vector of the features to be plotted
cont.plot	character of either "boxplot" or "violin"
cont.test	the significance test to be used for continuous variables. Use the name of the r basic tests.
disc.test	the significance test to be used for discrete variables. Should be "fisher","chi" or"both"
p.label, p.symnum.args	change the label style ("p.signif" or "p.format") of the significance. See stat_compare_means
cont.show.mean	boolean of whether to show the mean as a green dot for continuous variables
show.count	whether to show the total count of each catergory
cont.col	vector of colors for continuous variables
disc.col	list of named vectors for colors of discrete variables
disc.border, cont.border	show border or not for the discrete/continuous variables. pass to perc_barplot,violin
anno.textsize	change text size of the significance values
cont.anno.vjust, disc.anno.vjust	numeric to tweak the position of the pvalue annotation on the plot
aspect.ratio	the height width ratio of the figure
legend.position	position of the legends
highlight.signif, highlight.signif.col, signif.cutoff	Whether to highlight (highlight.signif) a significant plot with a color border (highlight.signif.col) according to the p-value cutoff (signif.cutoff). See miscPlots
plot.signif.only	whether to plot only the significant items
...	pass to theme for each individual plot

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