Programming with MATLAB

2013 The MathWorks, Inc. 1

Todays Objectives

Introduce you to programming with MATLAB

The MATLAB language Development tools

Show you how to program effectively in MATLAB

Demonstrate the range of programming styles supported

Interactive command line, scripts and functions, object-oriented programming
2

Demo: Assessment of Wind Turbine Locations

Goal: Create analysis to choose best wind turbine location by estimating power generation at multiple locations
Approach:
Interactively explore data and develop analysis approach

Load and preprocess data from observation towers Fit wind speed probability distribution Estimate average power

Automate to run on data from other locations

3

Calculating Average Turbine Power

( )

Wind Turbine Power Curve

Region I: > in = 0 Region II: in< < rated =
2 2 2 2

Region III: rated< < out

=

Region IV: > out

= 0
5

Demo Summary

Products Used
MATLAB Statistics Toolbox

Started working interactively

Import Tool Command line

Automated with a script

Command History Comments and sections Profiler to assess performance

Generalized with a function to handle multiple files

Range of Programming Techniques

value variable

structure

Refine and improve code

Maintainable Reusable/ more general Robust function script command line

Refine and Improve Code

Maintainable
Data types (structures, cell arrays, objects) Function types (subfunctions, nested functions,
private functions, packages)

Reusable / more general

Flexible calling syntax (varargin, varargout)

Robust
Error checking and validating inputs
(validateattributes, validatestring, error, inputParser)

Unit testing (matlab.unittest)

Development tools
Editor, Debugger, Code Analyzer, Profiler
8

Additional MATLAB Programming Language Features

Object-oriented (OO) programming with MATLAB

Objects combine related data and algorithms Class definition files describe object behavior Use matrix-based aspects of MATLAB with objects

Includes events and listeners, reference (handle) semantics, etc.

Parallel programming with Parallel Computing Toolbox

Speeds up MATLAB applications on multicore desktops, GPUs, and clusters Handles large datasets by distributing data across MATLAB workers Provides levels of parallel programming constructs

Simple constructs (built in support for Toolboxes, parfor, gpuArray) Advanced constructs (createJob/createTask, MPI, CUDA)

Sharing Results from MATLAB

Automatically generate reports

Create graphical user interfaces Package as an app Deploy applications to other environments Automatically generate standalone C code
10

Summary MATLAB for Programming

High-level language
Matrix-based Math and graphics functions Traditional programming language features

Interactive development environment

Tools, visualizations, and help

Supports a range of programming styles

Interactive command line, scripts and functions, object-oriented programming
11

Summary
The MATLAB Language

High-level language
Matrix-based Supports a range of programming styles No need for low-level administrative tasks

Math and graphics functions for engineering and science Includes features of traditional programming languages
Flow control and error handling Fundamental data types and advanced data structures
12

Summary
MATLAB Development Environment

Command Window
Provides immediate results by interactively executing commands

Data visualization
MATLAB Editor
Debugger Code Analyzer

MATLAB Profiler Additional tools for managing code and data

(e.g., shortcuts, code comparison, variable editor)
13

Summary
Multiple Ways to Get Help

doc

help <name>

Function Browser, function hints, tab completion

14

Resources

Webinars
Object-Oriented Programming in MATLAB MATLAB for C/C++ Programmers MATLAB to C Made Easy

Videos and code examples

MATLAB product page Documentation

MATLAB Central
Open exchange for the MATLAB and Simulink user community

15

MATLAB Central

File Exchange
Download free files Over 10K available including functions, apps, examples, and models

MATLAB Answers
Ask programming questions Search 18K+ existing answers

Cody
Challenge and expand your knowledge of MATLAB

Blogs
Read commentary from MathWorks engineers who design, build, and support MATLAB and Simulink
16

 2013 The MathWorks, Inc. 17