matrix_plot function

Graphical Tool for Visualizing Matrices

Plot of a matrix.

matrix_plot(x, ran = range(x, na.rm = TRUE), ylim = rev(c(0.5, nrow(x) + 0.5)),
            xlab = "Column", ylab = "Row",
            scales = list(alternating = c(1,1), tck = c(1,0),
                          x = list(at = pretty(1:ncol(x)), rot = 90),
                          y = list(at = pretty(1:nrow(x)))),
            at = NULL, colorkey = NULL, col = c("royalblue3", "white", "maroon3"),
            col.regions = NULL, ...)

Arguments

• x: matrix-like object.
• ran: range (can be used to enforce (-1,1), for example).
• ylim: y-axis limits in reverse order (for the rows to appear 'top down').
• xlab: x-axis label.
• ylab: y-axis label.
• scales: see levelplot(); if NULL, labels and ticks are omitted.
• at: see levelplot(). If NULL, a useful default is computed based on the given values in x.
• colorkey: see levelplot(). If NULL, a useful default is computed based on at.
• col: vector of length two (if all values of x are non-positive or all are non-negative; note that also a vector of length three is allowed in this case) or three (if x contains negative and positive values) providing the color key's default colors.
• col.regions: see levelplot(). If NULL, a useful default is computed based on at.
• ...: additional arguments passed to the underlying levelplot().

Returns

The plot, a Trellis object.

Details

Plot of a matrix.

Author(s)

Marius Hofert

Examples

## Generate a random correlation matrix
d <- 50
L <- diag(1:d)
set.seed(271)
L[lower.tri(L)] <- runif(choose(d,2)) # random Cholesky factor
Sigma <- L 
P <- cor(Sigma)

## Default
matrix_plot(P)
matrix_plot(P, ran = c(-1, 1)) # within (-1, 1)
matrix_plot(abs(P)) # if nonnegative
L. <- L
diag(L.) <- NA
matrix_plot(L.) # Cholesky factor without diagonal

## Default if nonpositive
matrix_plot(-abs(P))

## Changing colors
matrix_plot(P, ran = c(-1, 1),
            col.regions = grey(c(seq(0, 1, length.out = 100),
                                 seq(1, 0, length.out = 100))))

## An example with overlaid lines
library(lattice)
my_panel <- function(...) {
    panel.levelplot(...)
    panel.abline(h = c(10, 20), v = c(10, 20), lty = 2)
}
matrix_plot(P, panel = my_panel)