icorr: icorr

In raredd/iplotr: Interactive plotting functions

Description

Interactive correlation matrices (heat maps) with optional scatter plots.

Variables will be clustered (see details) and reordered by default. Cells will be labeled with the column names of the input data, and if a scatter plot is created, points will be labeled with the row names (this can be over-ridden by using the labels parameter).

Usage

icorr(
  mat,
  group,
  col,
  labels = NULL,
  cluster = TRUE,
  cor_method = "pearson",
  scatterplots = TRUE,
  plotOpts = NULL,
  digits = NULL
)

Arguments

mat	data matrix (observations x variables) of numeric values
group	vector of grouping (factor-like) variables for each observation
col	a vector of colors for each unique group of points in the scatter plots; note that colors will be assigned to the sorted levels of group
labels	optional character vector or named list of character vectors to label each point; if NULL, points will be labeled by index
cluster	logical or function; if TRUE, the variables will be clustered and reordered; if FALSE, no reordering will be done; otherwise, a custom clustering function may be given; see details
cor_method	character string indicating which correlation coefficient is to be computed; one of 'pearson' (default), 'kendall', or 'spearman': can be abbreviated; see cor
scatterplots	logical; if TRUE, scatter plots of the linked underlying data will be included
plotOpts	list of additional plot options; see icorrOpts
digits	integer indicating number of significant digits to use

Details

If col is given with no group variable, the colors for each observation will be recycled in order.

Default cluster method is stats::hclust(dist(x), method = 'average') which will return a list containing a named vector, "order", which is used to reorder the variables.

In order to pass a custom clustering function to cluster, the function must take a single input (a correlation matrix) and return either a vector or a list with a named vector, "order".

See Also

iplotCorr from the qtlcharts package

saveWidget

Other iplots: icurve(), idot(), iscatter(), itree()

Examples

## heatmap only
icorr(mtcars, scatterplots = FALSE)

## with scatter plots
icorr(mtcars, group = mtcars$cyl, col = c('blue','red','green'),
      plotOpts = list(corcolors = heat.colors(3)))

## larger matrix example
set.seed(1)
dat <- replicate(50, mtcars[, sample(1:11, 1), drop = FALSE])
dat <- do.call('cbind', dat)
icorr(dat, cluster = TRUE, group = mtcars$cyl)

## using a custom clustering function
icorr(dat, cluster = function(x) hclust(dist(x, method = 'maximum')))
icorr(dat, cluster = function(x) sample(seq.int(ncol(x))))
icorr(dat, cluster = function(x) list(order = order(rowMeans(x))))

raredd/iplotr documentation built on March 19, 2021, 12:45 a.m.