A heatmap is the most sweet way of visualizing the values of a matrix. The numbers are converted in a colored and intuitive graphic.
Let’s take the following matrix in CSV format, with hypothetical correlation values varying from 0 to 1.
Name,A,B,C,D,E var1,0.1,0.0,0.7,0.2,0.1 var2,0.2,0.3,0.4,1.0,0.0 var3,0.7,0.4,0.5,0.0,0.0 var4,0.3,0.1,0.5,0.3,0.6 vae5,0.0,0.8,0.9,1.0,0.5
Based only on the above matrix, it is hard to say what variables are more or less correlated. But in a heatmap the same matrix looks like this:
Here is the R code:
library("ggplot2") library("reshape2") library("plyr") library("scales") dataset <- read.csv("table.csv") dataset$Name <- with(dataset, reorder(Name, A)) dataset.m <- melt(dataset) dataset.m <- ddply(dataset.m, .(variable), transform, rescale = rescale(value)) (p <- ggplot(dataset.m, aes(variable, Name)) + geom_tile(aes(fill = rescale), colour = "white") + scale_fill_gradient(low = "saddlebrown", high = "gold")) base_size <- 10 p + theme_grey(base_size = base_size) + labs(x = "Title X axis", y = "Title Y axis") + scale_x_discrete(expand = c(0, 0)) + scale_y_discrete(expand = c(0, 0)) + theme(axis.ticks = element_blank(), axis.text.x = element_text(size = base_size * 0.8, angle = 330, hjust = 0, colour = "grey50")) ggsave("heatmap.png")