Assigning different color to NaN values in 2D matrix

I have a 2D matrix (600x500) which has different values ranging from 0 to 30 and some NaN values. I need to show it as a color image with each pixel having a code code corresponding to its value. I am doing this with imagesc/pcolor.

However, I need to show the NaN values in a completely different color, say white or black to differentiate them from rest of the elements in the matrix. How do I do this?

At present, the NaN values are getting interpreted as zeros and are being shown as the same color as zero.

[Sayan]

1 way that I figured out is to assign -1 to NaN values. The colour will be slightly differentiable from 0 but how to give more contrasting colour? e.g. 0 to 30 as blue to red and NaN as white.

[Matt Kindig]

I think your -1 trick is the right way to go. To force those values to be white (or black), you can define a custom colormap for your imagesc/pcolor call. For example:

mycolormap = [ ones(1,3); jet(30)];     %tack on a white row to your colormap
imagesc(...) %or pcolor(...)
colormap(mycolormap);  %call your custom colormap

[Cedric]

You could also overlay whatever you want to indicate these locations, e.g.

 imagesc(data) ;
 [r,c] = find(isnan(data)) ;
 hold on ; 
 plot(c, r, 'ro') ;

Answer 1

[Image Analyst]

Try this demo, make sure you read the comments. I make an array, then set random pixels to NaN. Then I set the nan pixels to red. By the way, don't use pcolor().

clc;  % Clear command window.
close all;  % Close all figure windows except those created by imtool.
workspace;  % Make sure the workspace panel is showing.
fontSize = 15;

image(imageArray);
% Apply a colormap
cmap = jet(256);
colormap(cmap);
colorbar
title('No NaNs', 'FontSize', fontSize);
% Enlarge figure to full screen.
set(gcf, 'units','normalized','outerposition',[0 0 1 1]);
% Give a name to the title bar.
set(gcf,'name','Demo by ImageAnalyst','numbertitle','off') 

% Now set random elements to NAN
uiwait(msgbox('Now set random elements to NAN'));
randomLocations = randperm(numel(imageArray), 2000);
imageArray(randomLocations) = nan;
image(imageArray);
colorbar
title('With 2000 NaNs', 'FontSize', fontSize);

% Now find the nans and set them to some gray level
% preferably one that isn't used at all,
% but let's set them to zero.
uiwait(msgbox('Now find the nans and set them to 0'));
imageArray(isnan(imageArray)) = 0;
image(imageArray);
colorbar
title('With NaNs set to zero', 'FontSize', fontSize);

uiwait(msgbox('Now find the 0s and set them to red'));
% Now let's change the color map for gray level 0
cmap(1, :) = [1 0 0]; % Set to red.
colormap(cmap); % Apply new colormap.