• Digital signal processing (DSP) – Analysis of data from motion sensors (accelerometers, gyroscopes, magnetometers), force plates, audio.  Also, we have considerable background in different filter types including IIR and FIR digital filters, zero-phase filtering, and Kalman filters.
• Statistics – We can analyze any set of data for correlation or regression equations.
• Interface development (GUIDE) for your functions
• Compiling your interfaces or functions for distribution
• General MATLAB development efficiency tools

Contact us for more information or a quote for your project from one of our consultants.

Notice (in the first example) that [1 3 7]*[1; 3; 7] could be written as [1 3 7]*[1 3 7]‘.  Notice in the third example that [1 3 7]*[1 3 7] could be written as [1 3 7]^2 which will also produce an error.

I think this these examples are fairly self-explanatory, but just understand your goals keeping what you learned in math class in mind. You can access MATLAB’s documentation for the element-by-element operator by typing “doc”, and searching for “arithmetic operators”.

1. In your interface, use a “panel” instead of an “axes”.
2. Set your panel’s tag name to “pnlMyPlot”
3. In your code, identify your plot target with
   subplot(1,1,1,'Parent',handles.pnlMyPlot)
4. Then, code your plot or subplot like normal. For example:
   subplot(2,1,1)
hold on
plot(rand(10,1))
ylabel('Y axis label')
xlabel('X axis label')
title('10 random numbers')
hold off

The attached example demonstrates using a panel instead of an axes for plotting in your interface. If you look at the code for each button, try uncommenting the code for a logical subplot using an axes, and check out what happens. Extract the two files into your MATLAB work directory. Subplots in GUIDE example

To make a shortcut, just right-click to the right of “Shortcuts” and choose “New Shortcut”. Give your shortcut a logical name (”Label”) and enter all the function calls or scripts in the callback.

The five most frequently used shortcuts in my workspace are:

1. Blank everything
   Label: “Blank”
   Callback:
   clear all
close all
clc
2. Clear variables
   Label: “Clear vars”
   Callback:
   clear all
3. Clear Command Window
   Label: “Clear CW”
   Callback:
   clc
4. Close all figures
   Label: “Close figures”
   Callback:
   close all
5. Stop debugging
   Label: “Stop debugging”
   Callback:
   dbquit

As of MATLAB R2007b, there is no quick way to create/enable different themes for your editor; however, we can still manually change settings. For some reason, I’m obsessed with dark backgrounds, light text, and simplistic fonts, so if you want to try different (better) settings, here are my personal settings.

To get your workspace similar to mine, follow these steps:

• Get the DejaVu font family
  • If you have Windows, get the latest package as a zip archive.
  • Open your “Fonts” folder (Start / Control Panel / Fonts)
  • Open your zip file of TrueType fonts, and find the “ttf” folder.
  • Copy everything in the “ttf” folder to your “Fonts” folder
• In MATLAB, go to File / Preferences…
  • Click “Fonts” on the left
  • Set your “Desktop code font” settings to DejaVu Sans Mono, Plain, 10 (code should be mono-spaced so everything aligns nicely)
  • Set your “Desktop text font” settings to DejaVu Sans, Plain, 10
• Stay in “Preferences”. Below “Fonts” open “Colors”.
  • Click the dropdown for “Text”; choose “More Colors…”; click “RGB” top right; make Red, Green, and Blue 240; click “OK”.
  • Click the dropdown for “Background”; choose “More Colors…”; click “RGB” top right; make Red, Green, and Blue 10; click “OK”.
  • Enable “M-Lint autofix highlight”; click the dropdown for it; choose “More Colors…”; click “RGB” top right; make Red, Green, and Blue 63; click “OK”.
• Stay in “Preferences”. In “Editor/Debugger”, choose “Display”.
  • Enable “Highlight current line”; select the dropdown for it; choose “More Colors…”; click “RGB” top right; make Red 10, Green 36, and Blue 106; click “OK”.
  • Enable “Highlight cells”; select the dropdown for it; choose “More Colors…”; click “RGB” top right; make Red, Green, and Blue 26; click “OK”.

data = cell(3,2);
data{1,1} = 'Data series 1';
data{1,2} = 'Data series 2';
data{2,1} = 1;
data{2,2} = 2;
data{3,1} = 3;
data{3,2} = 4;
data
xlswrite('myDataInExcel',data)

To learn how to use cell arrays properly, and understand the differences, type the following commands into your MATLAB window:

%% Make an array
myArray = zeros(3,1);
myArray(1) = pi;

%% Invalid assignment - can't put text in an array
myArray(2) = 'This is some text';

%% Another invalid assignment - can't put arrays in arrays
myArray(3) = rand(6,1);

%% Make a cell array
myCell = cell(3,1);
myCell{1} = pi;  % notice the curly braces
myCell{2} = 'This is some text';
myCell{3} = rand(6,1);
myCell  % display it

% accessing information from a cell
number_from_cell = myCell{1} % notice the curly braces
text_from_cell = myCell{2}
array_from_cell = myCell{3}

% notice the difference when you index the cell without curly braces:
number_from_cell = myCell(1)  % notice what happens when you use parenthesis
text_from_cell = myCell(2)
array_from_cell = myCell(3)

Remember that if you want your number in the cell array, but in text form, “sprintf” is really easy:

myCell{2} = sprintf('%g',pi) % quick example

I hope these examples demonstrate proper cell array creation and indexing.

1. Get newfun.m with supporting files and extract them to somewhere in your work directory
2. Modify “MSB_template.m” for your own name and company.

Next time you’re writing a new function (eg. myFunction which “Computes something”), start with this:

newfun('myFunction','Computes something')

Make sure to fill in a logical description, your inputs and outputs, make an example (if possible), and add some similar MATLAB functions to your “See also” section.

Here’s an example of one of my completed functions originally built with “newfun(’gen_sin’,'Make a sine wave’)”:

function output = gen_sin(f,fs,N)
%GEN_SIN  Make a sine wave
%   output = gen_sin(f,fs,N)
%
%   Description:
%     Create a frequency "f" of sampling rate "fs" of "N" many samples
%
%   Inputs:
%          f: frequency
%         fs: sampling rate
%          N: number of samples
%
%   Outputs:
%     output: generated frequency
%
%   Example
%     x = gen_sin(3,100,40); % make 3Hz with Fs of 100Hz, and 40 samples
%     plot(x);
%
%   See also GEN_NOISE

% Author: Simon
% Created: Dec 2005
% Copyright MSB Solutions 2005

ts = 1/fs;
t = ts*(0:N-1)';
output = sin(2*pi*f*t);

1. Get NSIS here: http://nsis.sourceforge.net/Main_Page
2. Get my example .nsi file (written out below)
3. Put my example .nsi file, your compiled .exe, your compiled .ctf, and the MCRInstaller X.X.exe in the same folder
4. Modify the example .nsi file for your application and your company name
5. Test it out. I recommend commenting out the section for the MCRInstaller using semicolons (;) at the beginning of each line for that section.
6. Right click on the .nsi file, and choose “Compile NSIS Script”. (You can do this manually by opening Start / All Programs / NSIS / MakeNSISW (Compiler GUI) – and following the instructions)
7. Run the generated installer, and check that everything works OK.
8. Uninstall your application
9. Uncomment the MCRInstaller section (remove the semicolons from the start of the line), and do #6 again.
10. Distribute your software however you want

Learning the NSIS scripting language can be tricky, but you essentially have all you need with my example. Hope this helps you out!

matlab_example_installer.nsi:

!define VERSION "1.0"
!define BUILD "1"

InstallDir C:\YourCompany\app\matlab_installer_example
Name "Installer Example ${VERSION} BUILD ${BUILD}"
OutFile "Installer Example V${VERSION}_${BUILD}.exe"

Section "application"
  CreateDirectory $INSTDIR
  SetOutPath $INSTDIR
  File matlab_installer_example.exe
  File matlab_installer_example.ctf
  CreateDirectory "$SMPROGRAMS\YourCompany"
  CreateShortCut "$SMPROGRAMS\YourCompany\matlab_installer_example.lnk" \
                 "$INSTDIR\matlab_installer_example.exe"
SectionEnd

; To try out the installation on your own
;computer, just comment out this section.
Section "MATLAB runtime 7.7"
  SetOutPath $INSTDIR
  File MCRInstaller.exe
  ExecWait MCRInstaller.exe
  Delete $INSTDIR\MCRInstaller.exe
SectionEnd

Section -AdditionalIcons
  CreateShortCut "$SMPROGRAMS\YourCompany\Uninstall.lnk" \
                 "$INSTDIR\uninstall.exe"
SectionEnd

Section -Post
  WriteUninstaller "$INSTDIR\uninstall.exe"
SectionEnd

Function un.onUninstSuccess
  HideWindow
  MessageBox MB_ICONINFORMATION|MB_OK "$(^Name) was \
             successfully removed from your computer."
FunctionEnd

Function un.onInit
  MessageBox MB_ICONQUESTION|MB_YESNO|MB_DEFBUTTON2 \
            "Are you sure you want to completely remove \
            $(^Name) and all of its components?" IDYES +2
  Abort
FunctionEnd

Section Uninstall
  Delete "$INSTDIR\uninstall.exe"
  Delete "$INSTDIR\matlab_installer_example.exe"
  Delete "$INSTDIR\matlab_installer_example.ctf"

  Delete "$SMPROGRAMS\YourCompany\matlab_installer_example.lnk"
  Delete "$SMPROGRAMS\YourCompany\Uninstall.lnk"

  RMDir "$INSTDIR"
  RMDir "$SMPROGRAMS\YourCompany"
SectionEnd

1. Writing math-based programming is really easy and quick in MATLAB
2. Building user interfaces to “wrap” your MATLAB code is very simple
3. Using MATLAB’s many different plots allows you display data in many ways
4. You can distribute your compiled applications freely without special licenses from MathWorks
5. Your m code is protected with 1024-bit RSA keys.

The only significant consequence for compiling MATLAB code is that you have to include MATLAB’s huge ~165MB runtime (found here: <MATLAB install directory>\toolbox\compiler\deploy\win32\MCRInstaller X.X.exe). Fortunately, you can install this just once on the target computer, and you can just update your compiled code as much as you want.

Here’s the compilation process for compiling a user interface built in MATLAB:

1. In MATLAB, type

   mbuild -setup

2. Choose your compiler (LCC is the default). If you have another C++ compiler, use that.
3. To make an executable for your user interface (example_file.fig, example_file.m, and other m files), type

   mcc -m example_file  % there are lots of options for mcc, type "doc mcc"

4. Take the compiled files (example_file.exe, example_file.ctf, and the MCRInstallerX.X.exe) and put them on the target computer
5. When you run example_file.exe the first time, it will extract the example_file.ctf (which contains all the required files to run your application except all your files are encrypted so no one can read your code)

For some applications, your users will have different settings, and you may want your application to remember the settings from the last time the application was opened by storing settings in an XML file.

xml_settings.m essentially works like this:

1. When the program is opened the first time, the default settings need to be written to the settings file -> we need to create the file and write settings.
2. When the program is opened after the file has been written, don’t overwrite the settings, just get them -> we need to open the file and read settings.
3. When the user has made a change to the settings, overwrite anything with the new setting -> we need to write or overwrite settings.

Here is the files and an example (extract to your work folder). Just type “saving_settings” in MATLAB.

xml_settings.m makes use of xml_write.m and xml_read.m by Jaroslaw Tuszynski found here.