How to edit the grid size in heatmap and how to make the color scale just a binary value?

I have a heat map as shown below. I want to make it look as continous plot as possible, or in other words I want to make it appear without individual grid borders. Also, is it possible to make a binary color seelection like red(high) and yellow(low). My data has a few inf values.Also, is there any way to remove the numbering from the x and y axis in the heatmap figure?

h  = heatmap(C);
colormap cool
% h.Colormap = cool;
h.ColorScaling = 'scaled';
h.ColorLimits = [0 10000];
h.GridVisible = 'off';
h.MissingDataColor = [0.8 0.8 0.8];
h.Title = 'Phase matrixs';
h.XLabel = '';
h.YLabel = '';
% caxis(h,[0 1]);

Answer 1

[Adam Danz]

"I have a heat map ... I want to make it look as continous plot as possible, ... I want to make it appear without individual grid borders."

Use imagesc(C) instead. You can smooth out the pixels using several different methods. For example, you could interpolate the values or, if you have the image processing toolbox, you can do a 2D gaussian filter as demonstracted below.

 imagesc(imgaussfilt(C,2))

"is there any way to remove the numbering from the x and y axis in the heatmap figure?"

Yes, but it's much easier to do if you use imagesc() by setting the xtick and ytick values.

"Also, is it possible to make a binary color seelection like red(high) and yellow(low). "

Yes, you can change the colormap

 set(gca, 'Colormap',[1 1 0; 1 0 0])

[Jay Vaidya]

Got it. Thank you!

[Jay Vaidya]

But I am not getting this working for other images. Like if you see below.

left(heatmap) and right(filtered image).

I have attached the C matrix below.

                % set(figure, 'Visible', 'off');
%                 h  = heatmap(C);
                h = imagesc(imgaussfilt(C,2));
                % colormap cool
                % h.Colormap = cool;
                % h.ColorScaling = 'scaled';
                % h.ColorLimits = [0 10000];
                % h.GridVisible = 'off';
                % h.MissingDataColor = [1 1 1];
                % h.Title = 'Phase matrixs';
                % h.XLabel = '';
                % h.YLabel = '';
                % caxis(h,[0 1]);
                set(gca, 'Colormap',[1 1 0; 1 0 0])

                set(gca,'xtick',[])
                set(gca,'xticklabel',[])
                set(gca,'ytick',[])
                set(gca,'yticklabel',[])

                %%
                dirname_phase_matrix = fullfile(dirname,sprintf('/figure%d_2.jpg',jaycp));
                saveas(h,dirname_phase_matrix)  % here you save the figure
                xlswrite(fullfile(dirname,sprintf('/matrix%d.xlsx',jaycp)),matrix);
                xlswrite(fullfile(dirname,sprintf('/C%d.xlsx',jaycp)),C);

[Adam Danz]

Here's what I get with those data and that code,

[Jay Vaidya]

Yes I got it. Some weird image saving issue. Thanks a lot!

Answer 2

[Image Analyst]

Why not use imshow()?

C = 1000 * rand(100);	% Create sample data.
C(10, 20) = inf;		% Put an inf in there, like the poster has.
imshow(C, [0, 1000], 'InitialMagnification', 800); % Display the image.
colormap(cool(256));	% Apply the colormap.
handleToColorBar = colorbar	% Display the colorbar.
% Turn off tick marks so that the colorbar won't have the numbers next to it.
handleToColorBar.Ticks = [];

I'm not sure what you mean by "binary color seelection like red(high) and yellow(low)". Does that mean that you only want two colors in there, not a gradient? If so, which value is the cutover value from red to yellow?

[Jay Vaidya]

This works good but can it be smoothened such that it looks like below (without grid borders?)

left(target) and right(yours)

[Image Analyst]

What are you doing to get C? The matrix in the workbook has some weird pattern where some pixels are 0 and others are 65535, and the "real" data is in some pattern. Doing this got rid of some bad rows and columns:

data = xlsread('C20.xlsx');
data(:, 1:4:end) = [];
data(4:4:end, :) = [];

[Jay Vaidya]

Thanks alot!