HeatMap not displaying correctly using ggplot()

Question

I am having a strange situation when I am trying to plot a heatmap on a dataset that I have which can be found here.

I am using the following code to plot the heat map:

xaxis<-c('density')
midrange<-range(red[,xaxis])
xaxis <- c(xaxis,'quality')
molten<-melt(red[,xaxis],'quality')

p <- ggplot(molten, aes(x = value, y = quality))
p <- p + geom_tile(aes(fill = value), colour = "white")
p <- p + theme_minimal()

# turn y-axis text 90 degrees (optional, saves space)
p <- p + theme(axis.text.y = element_text(angle = 90, hjust = 0.5))

# remove axis titles, tick marks, and grid
p <- p + theme(axis.title = element_blank())
p <- p + theme(axis.ticks = element_blank())
p <- p + theme(panel.grid = element_blank())

p <- p + scale_y_discrete(expand = c(0, 0)) 
# optionally remove row labels (not useful depending on molten)
p <- p + theme(axis.text.x = element_blank()) 

# get diverging color scale from colorbrewer
# #008837 is green, #7b3294 is purple
palette <- c("#008837", "#b7f7f4", "#b7f7f4", "#7b3294")


if(midrange[1] == midrange[2]) {
  # use a 3 color gradient instead
  p <- p + scale_fill_gradient2(low = palette[1], mid = palette[2], high = palette[4], midpoint = midrange[1]) +
    xlim(midrange[1],midrange[2])
}else{
  # use a 4 color gradient (with a swath of white in the middle)
  p <- p + scale_fill_gradientn(colours = palette, values = c(0, midrange[1], midrange[2], 1)) +
    xlim(midrange[1],midrange[2])
}
p

I am trying to plot the heat map on the variable Density and would like to use the variable quality as separation in my heat map. When I use the above code, I get the following plot:

It can be clearly seen that it is a blank image. This is happening because the range of the variable Density is very low, it doesn't happen if I change the variable to the one having a wider range (pH for example). Should ggplot automatically adjust to this? If not, how can I get ggplot to show the real plot?

Any help in this regard will be much appreciated.

Answer 1

So there are (at least) two problems here.

First, you have almost 1600 tiles in the x-direction, so specifying color="white" for the outline means that all you see is the outline, hence, white. Try taking this out.

Second, in your values=c(...) argument to scale_fill_gradientn(...) you seem to expect the midrange[1] and midrange[2] to be between (0,1), but midrange[2] = 1.003.

After taking out color="white" from the call to geom_tile(...), I get this: