How to increase color contrast for correlation plot in R-CodePudding

I have the following data:

dat <- structure(list(`2022.09.16` = c(1, 5, 1, 8, NA, 1, NA, 7, 4, 
NA), `2022.10.11` = c(2, 6, 4, 8, 1, 2, 2, 8, 7, 11), `2022.10.18` = c(4, 
8, 5, 8, 2, NA, 2, 7, 9, 9)), row.names = c("p1", "p2", "p3", 
"p4", "p5", "p6", "p7", "p8", "p9", 
"p10"), class = "data.frame")

That looks like this:

    2022.09.16 2022.10.11 2022.10.18
p1           1          2          4
p2           5          6          8
p3           1          4          5
p4           8          8          8
p5          NA          1          2
p6           1          2         NA
p7          NA          2          2
p8           7          8          7
p9           4          7          9
p10         NA         11          9

What I want to do is to make correlation plot based on column. I tried this:

GGally::ggcorr(dat,  label = TRUE )

I get this plot:

As you can see the contrast between low and high value is not very clear. What I'd like to get is 0.7 to be bluish. How can I achieve that?

CodePudding user response：

Use scale_fill_gradientn.

library(GGally)
ggcorr(dat,  label = TRUE )  
  scale_fill_gradientn(colors = c("#1b9e77","#d95f02","#7570b3"),
                       values = scales::rescale(c(-1, 0.7, 1)),
                       limits = c(-1, 1))

Adjust colors as desired.