ggpairs plot with heatmap of correlation values

Question

My question is twofold;

I have a ggpairs plot with the default upper = list(continuous = cor) and I would like to colour the tiles by correlation values

Answer 1

You can map a background colour to the cell by writing a quick custom function that can be passed directly to ggpairs. This involves calculating the correlation between the pairs of variables, and then matching to some user specified colour range.

my_fn <- function(data, mapping, method="p", use="pairwise", ...){

              # grab data
              x <- eval_data_col(data, mapping$x)
              y <- eval_data_col(data, mapping$y)

              # calculate correlation
              corr <- cor(x, y, method=method, use=use)

              # calculate colour based on correlation value
              # Here I have set a correlation of minus one to blue, 
              # zero to white, and one to red 
              # Change this to suit: possibly extend to add as an argument of `my_fn`
              colFn <- colorRampPalette(c("blue", "white", "red"), interpolate ='spline')
              fill <- colFn(100)[findInterval(corr, seq(-1, 1, length=100))]

              ggally_cor(data = data, mapping = mapping, ...) + 
                theme_void() +
                theme(panel.background = element_rect(fill=fill))
            }

Using the data in Marco's answer:

library(GGally)    # version: ‘1.4.0’

p1 <- ggpairs(sample_df, 
                   upper = list(continuous = my_fn),
                   lower = list(continuous = "smooth"))  

Which gives:

---

A followup question Change axis labels of a modified ggpairs plot (heatmap of correlation) noted that post plot updating of the theme resulted in the panel.background colours being removed. This can be fixed by removing the theme_void and removing the grid lines within the theme. i.e. change the relevant bit to (NOTE that this fix is not required for ggplot2 v3.3.0)

ggally_cor(data = data, mapping = mapping, ...) + 
           theme(panel.background = element_rect(fill=fill, colour=NA),
                 panel.grid.major = element_blank())