Skip to content

Matlab Quick Reference

Using waitbar

h = waitbar(0,'');
set(findall(h,'type','text'),'Interpreter','none');
for n=1:10
    waitbar(n/10, h, sprintf(Progress: %d/10,n) );
end
close(h)

Plot a data like a heat map

Use imagesc:

data = load('some_csv_file.csv');
imagesc(data);
colorbar;        % Add a colorbar to the plot
colormap('jet'); % Change the color map to "jet" type
caxis([-5 -2]);  % Set the range of values mapped by the colormap.
axis xy;         % change the direction of y-axis

isOctave = exist('OCTAVE_VERSION', 'builtin') ~= 0;
figHandle = gcf;
if(isOctave)
  print( sprintf('-f%d', figHandle.Number), '-djpg', print_fpath,  '-S800,600' );
else
  figHandle.PaperUnits = 'points';
  figHandle.PaperPosition = [0,0, figHandle.Position(3:4)]*3/4;
  figHandle.PaperSize = figHandle.Position(3:4)*3/4;
  figHandle.PaperPositionMode = 'manual';
  print( sprintf('-f%d', figHandle.Number), '-djpeg90', '-r0', print_fpath);
end

The flag -r0 sets the resolution to be the screen resolution. And configuring PaperPosition and PaperSize based on Position allows the output image size to be consistent with the image shown on the screen (here the figHandle.Units is assumed to be Pixels by default). A magic scaling factor of 3/4 is required in order to print an N-pixel image on screen to an N-pixel image in file. Whether this scaling factor is related to the screen resolution (after setting the -r0 flag) is to be investigated.

figHandles = findobj('Type','figure');
hall = findall(figHandles,'type','axes','tag','');
linkaxes(hall, 'x');

Use the criteria of empty 'tag' to filter out the axis handles that are the legends.

Set yscale to log in a Histogram Plot

axHandle = gca;

% A code snippet from MathWork to make the plot work with logscale on
% Y-axis
ph = get(axHandle,'children');
% Determine number of histogram patches
N_patches = length(ph);
for i = 1:N_patches
    % Get patch vertices
    vn = get(ph(i),'Vertices');
    % Adjust y location
    vn(:,2) = vn(:,2) + 1;
    % Reset data
    set(ph(i),'Vertices',vn)
end
set(axHandle,'yscale','log')

Create a Circulant Matrix from a Vector

To create a circulant matrix like the following:

\[ \left[ \begin{array}{ccccc} 1 & 16 & 8 & 4 & 2 \\ 2 & 1 & 16 & 8 & 4 \\ 4 & 2 & 1 & 16 & 8 \\ 8 & 4 & 2 & 1 & 16 \\ 16 & 8 & 4 & 2 & 1 \end{array} \right] \]

Run

x = [1 2 4  8 16];
a = gallery('circul', x);

Note: if x is an array, gallery() will first convert it to a vector (equivalent to x(:)).

Running from Perl and Keeping Perl Waiting until Matlab is done

system("matlab -nodisplay -nosplash -nodesktop -wait -r \"run myscript; exit\"");

Without the flags of -nodisplay -nosplash -nodesktop -wait, Perl runs Matlab in a forked process and will not wait for Matlab to finish its run.

Reference