Hmisc (version 5.1-3)

plotCorrM: plotCorrM

Description

Plot Correlation Matrix and Correlation vs. Time Gap

Usage

plotCorrM(
  r,
  what = c("plots", "data"),
  type = c("rectangle", "circle"),
  xlab = "",
  ylab = "",
  maxsize = 12,
  xangle = 0
)

Value

a list containing two ggplot2 objects if what='plots', or a data frame if what='data'

Arguments

r

correlation matrix

what

specifies whether to return plots or the data frame used in making the plots

type

specifies whether to use bottom-aligned rectangles (the default) or centered circles

xlab

x-axis label for correlation matrix

ylab

y-axis label for correlation matrix

maxsize

maximum circle size if type='circle'

xangle

angle for placing x-axis labels, defaulting to 0. Consider using xangle=45 when labels are long.

Author

Frank Harrell

Details

Constructs two ggplot2 graphics. The first is a half matrix of rectangles where the height of the rectangle is proportional to the absolute value of the correlation coefficient, with positive and negative coefficients shown in different colors. The second graphic is a variogram-like graph of correlation coefficients on the y-axis and absolute time gap on the x-axis, with a loess smoother added. The times are obtained from the correlation matrix's row and column names if these are numeric. If any names are not numeric, the times are taken as the integers 1, 2, 3, ... The two graphics are ggplotly-ready if you use plotly::ggplotly(..., tooltip='label').

Examples

set.seed(1)
r <- cor(matrix(rnorm(100), ncol=10))
g <- plotCorrM(r)
g[[1]]  # plot matrix
g[[2]]  # plot correlation vs gap time
# ggplotlyr(g[[2]])
# ggplotlyr uses ggplotly with tooltip='label' then removes
# txt: from hover text