mat2cols: Convert a numeric matrix to colours

In Keefe-Murphy/IMIFA: Infinite Mixtures of Infinite Factor Analysers and Related Models

Convert a numeric matrix to colours

Description

Converts a matrix to a hex colour code representation for plotting using plot_cols. Used internally by plot.Results_IMIFA for plotting posterior mean loadings heatmaps.

Usage

mat2cols(mat,
         cols = NULL,
         compare = FALSE,
         byrank = FALSE,
         breaks = NULL,
         na.col = "#808080FF",
         transparency = 1,
         ...)

Arguments

mat	Either a matrix or, when compare is TRUE, a list of matrices.
cols	The colour palette to be used. The default palette uses viridis. Will be checked for validity by is.cols.
compare	Logical switch used when desiring comparable colour representations (usually for comparable heat maps) across multiple matrices. Ensures plots will be calibrated to a common colour scale so that, for instance, the colour on the heat map of an entry valued at 0.7 in Matrix A corresponds exactly to the colour of a similar value in Matrix B. When TRUE, mat must be supplied as a list of matrices, which must have either the same number of rows, or the same number of columns.
byrank	Logical indicating whether to convert the matrix itself or the sample ranks of the values therein. Defaults to FALSE.
breaks	Number of gradations in colour to use. Defaults to length(cols). Alternatively, a vector of breakpoints for use with cut.
na.col	Colour to be used to represent missing data. Will be checked for validity by is.cols.
transparency	A factor in [0, 1] modifying the opacity for overplotted lines. Defaults to 1 (i.e. no transparency). Only relevant when cols is not supplied, otherwise the supplied cols must already be adjusted for transparency.
...	Catches unused arguments.

Value

A matrix of hex colour code representations, or a list of such matrices when compare is TRUE.

See Also

plot_cols, heat_legend, is.cols, cut

Examples

# Generate a colour matrix using mat2cols()
mat      <- matrix(rnorm(100), nrow=10, ncol=10)
mat[2,3] <- NA
cols     <- heat.colors(12)[12:1]
(matcol  <- mat2cols(mat, cols=cols))

# Use plot_cols() to visualise the colours matrix
par(mar=c(5.1, 4.1, 4.1, 3.1))
plot_cols(matcol)

# Add a legend using heat_legend()
heat_legend(mat, cols=cols); box(lwd=2)

# Try comparing heat maps of multiple matrices
mat1     <- cbind(matrix(rnorm(100, sd=c(4,2)),   nr=50, nc=2, byrow=TRUE), 0.1)
mat2     <- cbind(matrix(rnorm(150, sd=c(7,5,3)), nr=50, nc=3, byrow=TRUE), 0.1)
mat3     <- cbind(matrix(rnorm(50,  sd=1),        nr=50, nc=1, byrow=TRUE), 0.1)
mats     <- list(mat1, mat2, mat3)
colmats  <- mat2cols(mats, cols=cols, compare=TRUE)
par(mfrow=c(2, 3), mar=c(1, 2, 1, 2))

# Use common palettes (top row)
plot_cols(colmats[[1]]); heat_legend(range(mats), cols=cols); box(lwd=2)
plot_cols(colmats[[2]]); heat_legend(range(mats), cols=cols); box(lwd=2)
plot_cols(colmats[[3]]); heat_legend(range(mats), cols=cols); box(lwd=2)

# Use uncommon palettes (bottom row)
plot_cols(mat2cols(mat1, cols=cols)); heat_legend(range(mat1), cols=cols); box(lwd=2)
plot_cols(mat2cols(mat2, cols=cols)); heat_legend(range(mat2), cols=cols); box(lwd=2)
plot_cols(mat2cols(mat3, cols=cols)); heat_legend(range(mat3), cols=cols); box(lwd=2)

Keefe-Murphy/IMIFA documentation built on Jan. 31, 2024, 2:15 p.m.