How to plot different and independent color gradient legend in heatmap.2 (R)

I obtain a matrix from calculation where the values are between 0 and 4. I use the function heatmap.2 to plot.

The idea is to have a color code and with different gradient in function of the cell value:

from 0 to 0.99: gradient of red

from 1 to 1.99: gradient of yellow

from 2 to 2.99: gradient of green

from 3 to 4: gradient of grey

As you can see, the gradients are not independent, the colors smoothly change from a range to another (by example, I got orange from red to yellow). How to obtain 4 different and independent gradient? Mean, from 0 to 0.99 only red gradient and from 1 to 1.99 only yellow gradient etc

Here the code I used (I create a random matrix for the example):

library(gplots)

# create a matrix of random values from [0, 4]
random.matrix  <- matrix(runif(100, min = 0, max = 4), nrow = 10)

quantile.range <- quantile(random.matrix, probs = seq(0, 1))
palette.breaks <- seq(quantile.range["0%"], quantile.range["100%"], 0.25)
color.palette  <- colorRampPalette(c("red", "yellow", "green", "grey"))(length(palette.breaks)-1)
heatmap.2(
  random.matrix,
  dendrogram = "none",
  scale      = "none",
  trace      = "none",
  symbreaks=FALSE,
  symkey=FALSE, 
  key        = TRUE,
  key.title = 'class information',
  key.xlab = 'class number',
  keysize = 2,
  #( "bottom.margin", "left.margin", "top.margin", "right.margin" )
  key.par=list(mar=c(3,3.5,2,0)),
  labRow     = NA,
  labCol     = NA,
  col    = color.palette,
  breaks = palette.breaks,
)
legend(x="topright", legend=c("class 1: city", "class 2: crops", "class 3: forest", "class 4: road"),fill=c("red", "yellow", "green", "grey", ""))

Here the plot:

I found this way:

# gplots contains the heatmap.2 function
library(gplots)

# create a matrix of random values from [0, 4]
random.matrix  <- matrix(runif(100, min = 0, max = 4), nrow = 10)

quantile.range <- quantile(random.matrix, probs = seq(0, 1))
palette.breaks <- seq(quantile.range["0%"], quantile.range["100%"], 0.25)

colfunc1 <- colorRampPalette(c("white", "red"))
colfunc2 <- colorRampPalette(c("white", "yellow"))
colfunc3 <- colorRampPalette(c("white", "green"))
colfunc4 <- colorRampPalette(c("grey", "black"))

color.palette  <- colorRampPalette(c(colfunc1(4), colfunc2(4), colfunc3(4), colfunc4(3)))(length(palette.breaks)-1)

heatmap.2(
  random.matrix,
  dendrogram = "none",
  scale      = "none",
  trace      = "none",
  symbreaks=FALSE,
  symkey=FALSE, 
  key        = TRUE,
  key.title = 'class information',
  key.xlab = 'class number',
  keysize = 2,
  #( "bottom.margin", "left.margin", "top.margin", "right.margin" )
  key.par=list(mar=c(3,3.5,2,0)),
  labRow     = NA,
  labCol     = NA,
  col    = color.palette,
  breaks = palette.breaks,
)
legend(x="topright", legend=c("class 1: city", "class 2: crops", "class 3: forest", "class 4: road"),fill=c("red", "yellow", "green", "grey", ""))