File Exchange

image thumbnail

Color Palette Tables (.cpt) for Matlab

version (1.43 MB) by Kelly Kearney
Create and apply GMT-style colormaps in Matlab


Updated 20 Aug 2016

GitHub view license on GitHub

cptcmap.m: color palette tables (.cpt) for Matlab

Comments and Ratings (25)

Liang Han

Lei Cheng

Thank you so much.! It is awesome!


helpful, thank you

Great tool!
Instead I can't figure out how to obtain shading interp. Matlab continues to warning about it. Any ideas please?

E. Cheynet

Chad Greene

Chad Greene

Superb. Thanks for writing this easy-to-use, well-documented function.

I've addressed Seth's question via email. But in case others encounter this, the behavior he was seeing is as intended. While the length of the resulting colormap can be specified via the ncol input, the number of unique colors in that colormap is determined by the .cpt file itself. The GMT_Haxby colormap defines 32 specific colors, so increasing the resolution of the colormap via ncol will create a larger colormap array, but many of these values will simply be replicates.


More detail on my comment below:

cptcmap('GMT_Haxby','ncol', 4) produces a contour plot with four colors;
cptcmap('GMT_Haxby','ncol', 8) produces a contour plot with eight colors;
cptcmap('GMT_Haxby','ncol', 32) produces a contour plot with thirty-two colors;

However, cptcmap('GMT_Haxby','ncol', 256) produces a contour plot with 32 colors, not 256 colors. Any number above 32 leads to a plot with just 32 colors.

Matlab novice, hope I'm not missing something trivial!


I'm having trouble getting these contour plots to use more than ~32 colors with this function.

If I use the command:


then contour plots plot using the jet colorscheme with 256 colors. However, if I use the command:

cptcmap('GMT_Haxby','ncol', 256);

then all of the subsequent contour plots display with only ~32 colors.

Any suggestions? Thanks!


Function works brilliantly!! Thanks so much!


Sorry, I always forget that roundn is part of the Mapping toolbox. Just replace the roundn call:

space = roundn(space, -3);


space = round(space*1e3)/1e3;


I don't have access to the "roundn" function. Is there a work around so that I can get your function working?

Frank Engel

Excellent and very useful. Works as expected, and had no problems using with gradients provided from cpt-city (

Very useful for bathymetric mapping. Please note that there is a dependency on the mapping toolbox due to the use of the 'roundn' function. This could easily be replaced by 1 line of standard Matlab code.


@camilo melo,

I don't have R20011b, so I can't test it, but I can't think of any reason it wouldn't work in that version. Based on the error message you mention, my guess is you tried to call the function without any input arguments; it needs at least one (the name of the colormap).

camilo melo

hi, does this function works for MATLAB R2011b?? i tried to use it but got this error message Index exceeds matrix dimensions.

Error in cptcmap (line 123)
[blah, blah, ext] = fileparts(varargin{1});

Another one, trying


Just had to replace

if ~strcmp(varargin{1}(end-3:end), '.cpt') % without extension


if (numel(varargin{1}) <= 4) || ~strcmp(varargin{1}(end-3:end), '.cpt') % without extension

Another one, trying


Just had to replace

if ~strcmp(varargin{1}(end-3:end), '.cpt') % without extension


if (numel(varargin{1}) <= 4) || ~strcmp(varargin{1}(minend-3:end), '.cpt') % without extension

You might want to replace l.83:

cptpath = './cptfiles';


cptpath = fullfile(fileparts(which('cptcmap')), 'cptfiles');

That way the source code needs not be modified when first downloading the package or later when moving the routine files around.

I was having problems trying


where elevation.cpt was downloaded from cpt-city and placed in the cptpath; it turns out that l.98,

if exist(varargin{1}, 'file') % full filename and path given

would find this other file instead:

which elevation
C:\Program Files\MATLAB\R2009b\toolbox\map\map\elevation.m

You might want to replace the block l.98-111 with the following (minimally modified):

if ~strcmp(varargin{1}(end-3:end), '.cpt') % without extension
varargin{1} = [varargin{1} '.cpt'];
if exist(varargin{1}, 'file') % full filename and path given
filename = varargin{1};
else % only file name given
filename = fullfile(cptpath, varargin{1});
if ~exist(filename, 'file')
error('Specified .cpt file not found');

Finally, not sure why the __MACOSX subdirectory is being distributed.

Great idea, always wanted to make my matlab maps look like gmt's. Also thanks for pointing out to cpt-city, it's a great collection.

You might want to use the acknowledgement field in the FEX form, so that your submission is back-linked from related entries, e.g., 28982 and the ones linked from there.


If by the inverse colormap, you mean you want to flip the direction of the colors, then yes, use the 'flip' option:

cptcmap('GMT_seis', 'flip', true)


Hi. Very nice package. Is it possible to use the inverse colormap of GMT_seis.cpt?


- Linked to GitHub repository
- Added cptcbar.m to create colorbars
- Now allows for multiple folders holding .cpt files

Corrected bug in colormap length calculation

- Corrected bug in HSV color calculation
- Updated for 2014b graphics

Added error check in case no input is supplied. Also corrected the roundn issue (correctly, this time).

Replaced roundn call with manual calculation to remove unintended dependency on Mapping Toolbox.

-bug fixes where some input filenames were misinterpreted (thanks to Felipe)
-colormap display option no longer relies on external functions, has cleaner display, and renders more quickly

MATLAB Release Compatibility
Created with R2010a
Compatible with any release
Platform Compatibility
Windows macOS Linux

Discover Live Editor

Create scripts with code, output, and formatted text in a single executable document.

Learn About Live Editor