How to add p values for correlation coefficients p

Question

I am making a scatterplot matrix using lattice and plotting the correlation coefficients of 12 variables in the upper half of the panel. I would also like to add the p values beneath the correlation coeffiecients or stars indicating their level of significance. Here is my R code. How can I achieve this? Many thanks in advance!

Here is a sample of my data

d.corr1 = structure(list(maxt1.res = c(-0.944678376630112, 0.324463929632583, 
-1.18820341118942, -0.656600399095673, 0.332432965913295, 0.696656683837386
), maxt2.res = c(1.81878373188327, -0.437581385609662, 0.305933316224282, 
-3.20946216261864, 0.629812177862245, -1.49044366233353), maxt3.res = c(-1.21422295698813, 
-1.31516252550763, 0.570370111383564, 1.73177495368256, 2.18742200139099, 
0.413531254505875), mint1.res = c(0.783488332204165, 0.35387082927864, 
-0.528584845400234, 0.772682308165534, 0.421127289975828, 1.06059010003109
), mint2.res = c(0.262876147753049, 0.588802881606123, 0.745673830291112, 
-1.22383100619312, -1.01594162784602, -0.135018034667641), mint3.res = c(0.283732674541107, 
-0.406567031719476, 0.390198644741853, 0.860359703924238, 1.27865614582901, 
0.346477970454206), sr1.res = c(1.7258974480523, -1.71718783477085, 
3.98573602228491, -4.42153098079411, 0.602511156003456, -3.07683756735513
), sr2.res = c(9.98631829246284, -6.91757809846195, 0.418977023594041, 
-6.10811634134865, 14.6495418067316, 2.44365146778955), sr3.res = c(-3.8809447886743, 
2.35230122374257, 2.8673756880306, 7.1449786041902, 2.07480997224678, 
4.93316979213985), rain1.res = c(0.112986181584307, 0.0445969189874017, 
-0.446757191502526, 1.76152475011467, -0.395540856161192, -0.175756810329735
), rain2.res = c(-0.645121126413379, 1.74415111794381, -0.122876137090066, 
1.68048850848576, -0.570490345329031, 0.00308540146622738), rain3.res = c(-0.202762644577954, 
0.0528174267822909, -0.0616752465852931, -0.167769364680304, 
-0.152822027502996, -0.139253335052929)), .Names = c("maxt1.res", 
"maxt2.res", "maxt3.res", "mint1.res", "mint2.res", "mint3.res", 
"sr1.res", "sr2.res", "sr3.res", "rain1.res", "rain2.res", "rain3.res"
), row.names = c(NA, 6L), class = "data.frame")


attach(d.corr1)
library(lattice)
library(RColorBrewer)
splom(~d.corr1[seq(1:12)], lower.panel = panel.splom,
  upper.panel = function(x, y, ...) {
      panel.fill(col = brewer.pal(9, "RdBu")[ round(cor(x, y) * 4 + 5)])
      cpl <- current.panel.limits()
      panel.text(mean(cpl$xlim), mean(cpl$ylim), round(cor(x, y),2), font=2)
  },

 scales = list(x = list( draw = TRUE, cex=0.1)), type = c("g", "p", "smooth"),layout =           c(1, 1), pscales=0, pch=".",
 main="correlation between the weather variables after removing district F.E and yearly trends")
dev.off()

detach(d.corr1)

Answer 1

Base graphics solution for your question is given below.

panel.cor <- function(x, y, digits = 2, cex.cor, ...)
{
  usr <- par("usr"); on.exit(par(usr))
  par(usr = c(0, 1, 0, 1))
  # correlation coefficient
  r <- cor(x, y)
  txt <- format(c(r, 0.123456789), digits = digits)[1]
  txt <- paste("r= ", txt, sep = "")
  text(0.5, 0.6, txt)

  # p-value calculation
  p <- cor.test(x, y)$p.value
  txt2 <- format(c(p, 0.123456789), digits = digits)[1]
  txt2 <- paste("p= ", txt2, sep = "")
  if(p<0.01) txt2 <- paste("p= ", "<0.01", sep = "")
  text(0.5, 0.4, txt2)
}

pairs(iris, upper.panel = panel.cor)

I made this by modifying example provide for `pairs' function.

Answer 2

Another option is to use panel.text twice , with different adj parameter.

For example :

splom(~d.corr1[seq(1:12)], lower.panel = panel.splom,
      upper.panel = function(x, y, ...) {
        panel.fill(col = brewer.pal(9, "RdBu")[ round(cor(x, y) * 4 + 5)])
        cpl <- current.panel.limits()
        ## translate upward 
        panel.text(mean(cpl$xlim), mean(cpl$ylim), round(cor(x, y),2), font=2,
                   adj=c(0.5,-0.6))
        ## translate downward
        panel.text(mean(cpl$xlim), mean(cpl$ylim), round( cor.test(x,y)$p.value, 2), font=1,
                   adj=c(0.5,0.6),col='blue')
      },

Answer 3

Given that you have offered no data, I will assume you plan to do these calculations external to the plot. Let's assume your p-values are in a vector named p_vals. The instead of round(cor(x, y),2) as the third argument to text, use:

paste( round(cor(x, y),2), "\n", p_vals)

With data you could do it all within lattice using the same strategy:

 paste( round(cor(x, y),2) ,"\n", round( cor.test(x,y)$p.value, 2) )