ggcorrplot draws a correlation matrix as a
ggplot2 plot. Because the result is a plain ggplot
object, you can restyle it, annotate it, and combine it with other
layers using the usual + syntax.
It can:
cor_pmat().Learn more in ggcorrplot: Correlation Matrix Heatmap in R with ggplot2.
Install the released version from CRAN:
install.packages("ggcorrplot")Or the development version from GitHub:
if (!require(devtools)) install.packages("devtools")
devtools::install_github("kassambara/ggcorrplot")library(ggcorrplot)The examples below use the mtcars data set.
cor() builds the correlation matrix and
cor_pmat() [in ggcorrplot] computes the
matrix of correlation p-values.
data(mtcars)
corr <- round(cor(mtcars), 1)
corr[1:4, 1:4]
#> mpg cyl disp hp
#> mpg 1.0 -0.9 -0.8 -0.8
#> cyl -0.9 1.0 0.9 0.8
#> disp -0.8 0.9 1.0 0.8
#> hp -0.8 0.8 0.8 1.0
# Matrix of correlation p-values
p.mat <- cor_pmat(mtcars)
p.mat[1:4, 1:4]
#> mpg cyl disp hp
#> mpg 0.000000e+00 6.112687e-10 9.380327e-10 1.787835e-07
#> cyl 6.112687e-10 0.000000e+00 1.802838e-12 3.477861e-09
#> disp 9.380327e-10 1.802838e-12 0.000000e+00 7.142679e-08
#> hp 1.787835e-07 3.477861e-09 7.142679e-08 0.000000e+00The default draws each correlation as a colored square;
method = "circle" encodes the value with the circle area
instead.
ggcorrplot(corr)
ggcorrplot(corr, method = "circle")
scale.square = TRUE sizes the squares by the absolute
correlation, so strong correlations dominate;
cell.grid = TRUE draws a light box around every cell so the
glyphs sit inside a grid instead of floating on the axis lines.
ggcorrplot(corr, scale.square = TRUE, cell.grid = TRUE, outline.color = "white")
ggcorrplot(corr, method = "circle", cell.grid = TRUE)
hc.order = TRUE reorders the variables by hierarchical
clustering so that correlated variables sit together.
hc.rect then draws rectangles around the clusters obtained
by cutting the tree.
ggcorrplot(corr, hc.order = TRUE, outline.color = "white")
ggcorrplot(corr, hc.order = TRUE, hc.rect = 3, outline.color = "white")
For a symmetric matrix the two triangles are redundant, so you can keep just one.
ggcorrplot(corr, hc.order = TRUE, type = "lower", outline.color = "white")
ggcorrplot(corr, hc.order = TRUE, type = "upper", outline.color = "white")
lower.method and upper.method draw a
different glyph in each triangle — here the
coefficients as numbers below the diagonal and circles above it, with
the variable names on the diagonal. Adding cell.grid = TRUE
boxes every cell for the tidy corrplot look.
ggcorrplot(corr,
lower.method = "number", upper.method = "circle",
cell.grid = TRUE, show.legend = FALSE
)
ggcorrplot(corr, hc.order = TRUE, type = "lower", lab = TRUE)
Passing p.mat marks the cells whose correlation is not
significant at sig.level (default 0.05). By default a cross
is drawn over them (insig = "pch");
insig = "blank" hides them instead.
# Cross out the non-significant coefficients
ggcorrplot(corr, hc.order = TRUE, type = "lower", p.mat = p.mat)
# Leave them blank
ggcorrplot(corr, hc.order = TRUE, type = "lower", p.mat = p.mat, insig = "blank")
insig = "stars" flips the emphasis: instead of crossing
out the non-significant cells, it marks the significant
ones with significance stars (***, **,
* for p <= 0.001, 0.01, 0.05 – fixed thresholds, not
sig.level). With the default lab = FALSE this
is a standalone significance map; with lab = TRUE the stars
are appended to the coefficients (e.g. -0.85***).
ggcorrplot(corr, p.mat = p.mat, insig = "stars")
ggcorrplot() returns a ggplot object, so any ggplot2
theme applies. colors sets the low / mid / high gradient.
For the ggplot2 side of this, see ggplot2
Colours in R: Change Colours by Group and ggplot2
Themes in R: Customize the Look.
ggcorrplot(corr,
hc.order = TRUE, type = "lower", outline.color = "white",
ggtheme = ggplot2::theme_gray,
colors = c("#6D9EC1", "white", "#E46726")
)