1

Given a matrix like:

m <- matrix(ncol = 4, nrow = 3)
colnames(m) <-c("attribute1-1","attribute1-2","attrbute1-3","attribute2")
rownames(m) <-c("sample1","sample2","sample3")
m[1,] <- c(5,7,0,2)
m[2,] <- c(0,0,0,6)
m[3,] <- c(0,0,4,1)

I would like to make a heat map where attributes 1-1 : 1-3 are used to indicate presence/absence (whether that individual entry is blank or not, depending if there is a positive value or 0) and attribute 2 is used to indicate intensity. In other words: if, according to the attribute1 column there is the presence of that row in that instance of attribute 1 (e.g. positive value for attribute1-2 for row1) then the intensity is prescribed by attribute 2.

I think the presence/absence could be done using a formula such as "decostand": X.pa <- decostand(X,method="pa")

And heatmaps can be created from several packages, notably "Superheat".

I can't seem to get anything working, though. Any advice as to how much to continue would be much appreciated. Thank you in advance.

kalgarianer
  • 15
  • 4
  • Please provide a reproducible example of your dataset by following this code: https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example – dc37 Apr 04 '20 at 03:11

2 Answers2

1

Thank you for adding sample data. Solution using ggplot2:

m <- matrix(ncol = 4, nrow = 3)
colnames(m) <-c("attribute1-1","attribute1-2","attrbute1-3","attribute2")
rownames(m) <-c("sample1","sample2","sample3")
m[1,] <- c(5,7,0,2)
m[2,] <- c(0,0,0,6)
m[3,] <- c(0,0,4,1)

m
#>         attribute1-1 attribute1-2 attrbute1-3 attribute2
#> sample1            5            7           0          2
#> sample2            0            0           0          6
#> sample3            0            0           4          1


# library
library(tidyverse)


# adjust data
dat <- m %>%
  as_tibble(rownames="sample") %>%
  gather(name,value, -c(sample,attribute2)) %>% 
  mutate(attribute2=case_when(value==0 ~ 0, TRUE ~attribute2)) 

#make plot
ggplot(dat,aes(name,sample)) +
  geom_tile(aes(fill = attribute2), colour = "white") +
  scale_fill_gradient(low = "white", high = "steelblue")

Created on 2020-04-04 by the reprex package (v0.3.0)

captcoma
  • 1,768
  • 13
  • 29
1

superheat https://rlbarter.github.io/superheat/index.html is a great library to quickly visualize matrices

library(superheat)
m <- matrix(ncol = 4, nrow = 3)
colnames(m) <-c("attribute1-1","attribute1-2","attrbute1-3","attribute2")
rownames(m) <-c("sample1","sample2","sample3")
m[1,] <- c(5,7,0,2)
m[2,] <- c(0,0,0,6)
m[3,] <- c(0,0,4,1)

superheat(m)

enter image description here

bioSlayer
  • 85
  • 5