Add color bars beside rotated triangular matrix-CodePudding

I wish to create an triangular matrix, rotate it 45 degrees clockwise, and attach color bars on the sides of the rotated triangular matrix. Here is my code:

# The data
library(tidyverse)

x <- 1:10
y <- 1:10
data <- expand.grid(X=x, Y=y)
data$X <- as.numeric(data$X)
data$Y <- as.numeric(data$Y)
data$Z <- 1:(10*10)

# Create upper triangular matrix
zz <- t(matrix(data$Z, 10, 10))
zz[lower.tri(zz)] <- NA
data$zz <- c(t(zz))

# Use "-Y" so that the first row of data is plotted in the first row
p1 <- ggplot(data, aes(X, -Y, fill= zz))   
  geom_tile()  
  # scale_fill_gradient(low="blue", high="red")  
  scale_fill_gradient(low = "#132B43", high = "#56B1F7", space = "Lab", na.value="white")  
  theme_bw()  
  theme(axis.title = element_blank(),
        axis.text = element_blank(),
        axis.ticks = element_blank(),
        panel.grid.major = element_blank(),
        panel.grid.minor = element_blank(),
        panel.border = element_blank(),
        panel.background = element_blank())

This is p1, which looks fine:

I wish to rotate this matrix 45 degrees clockwide. I did so following the suggested code

Here is what I want:

Since I wish to rotate p1 45 degrees clockwise, the diagonal should be verticle and there should not be missing value inside the triangular matrix. How to do the rotation?

Additionally, I want to add color bars by sides of the heatmap, as shown by the pink circled color bars in the figure below:

My data has 10 columns and 10 rows, I wish to plot color bars like above for my own heatmap using 5 different colors, with each color spanning two adjacent rows/columns. How to add?

CodePudding user response：

zz %>%
  as.data.frame() %>%
  mutate(row = 11 - row_number()) %>%
  pivot_longer(-row, names_to = "col") %>%
  filter(!is.na(value)) %>%
  # slice(1:12) %>%
  mutate(col = parse_number(col),
         col_new = col - 5 - (row col)/2,
         row_new = row - (10 - col)) %>% 
  ggplot(aes(col_new, row_new, fill = value))  
  geom_tile(width = 1)

CodePudding user response：

p1_fmt <- p1   guides(fill = "none")   coord_equal()
tri_file <- tempfile()
ggsave(tri_file, p1_fmt, width = 3, height = 3, device = "png")

p1_guide <- cowplot::get_legend(p1)
guide_file <- tempfile()
ggsave(guide_file, p1_guide, width = 1, height = 1.5, device = "png")


library(magick)
tri <- image_read(tri_file) %>%
  image_rotate(45) %>%
  image_trim()
guide <- image_read(guide_file)

image_append(c(tri, guide))