pcaServer: Shiny Module – PCA plots

In bihealth/bioshmods: What the Package Does (One Line, Title Case)

Shiny Module – PCA plots

Description

Shiny Module – PCA plots

Usage

pcaUI(id, datasets = NULL)

pcaServer(
  id,
  pca,
  covar,
  idcol = "ID",
  threeD = FALSE,
  colorBy = NULL,
  symbolBy = NULL
)

Arguments

id	identifier of the shiny module (character vector)
datasets	if not NULL, a character vector specifying the data sets (see Details)
pca	pca matrix – columns correspond to principal components, rows to observations. Rows must be named and must correspond to the ID column of the covariate data frame.
covar	data frame containing covariates. The identifiers of the samples in the covariate data frame are taken from the ID column (by default, "ID").
idcol	name of the ID column in the covariate data frame.
threeD	whether the plot should be three-dimensional by default
colorBy	selected covariate to use for coloring the plot
symbolBy	selected covariate to use for symbols on the plot

Datasets

Rather than specifying pca as a matrix and covariates as a data frame, these arguments can be, respectively, named lists of matrices and of data frames. This allows to switch between different data sets (e.g. with different samples, covariates) or between different representations (e.g. different PCA variants or other transformations such as Umap). Note that if the pca argument is a list, then the covar argument must be a list, too, and all the names of the pca list must be also in the covar list.

Examples

if(interactive()) {
  data(iris)
  covar <- iris
 
  pca <- prcomp(iris[,1:4], scale.=TRUE)
 
  ui <- fluidPage(pcaUI("pca"))
 
  server <- function(input, output, session) {
    pcaServer("pca", pca$x, covar, colorBy="Species")
  }
 
  shinyApp(ui, server)
}

bihealth/bioshmods documentation built on July 1, 2023, 4:32 a.m.