Programming Basics

Introductory resources to learn programming

There are some very good free on-line tutorials out there to help you begin to learn the basics of pretty much any computer language.

• A nice introduction to Matlab programming can be found here.

• Mathworks, the company that provides Matlab, has an extensive set of help and reference pages. See this page to start. Of these introductory modules, psychologists will likely find the following most helpful:

  • Getting Started

  • Language Fundamentals

  • Data Import and Analysis

  • Programming Scripts and Functions

• Getting a comprehensive book for learning can also be helpful. There are many, many commands and functions that have already been written and shared that can make your life easier, but remembering their details is not always easy.

  • I learned using a book called Mastering MATLAB by Hanselman & Littlefield. It's older now, but it holds up!

  • Matlab: A practical introduction to programming and problem solving (Attaway) isn't as comprehensive but may be an easier introduction for complete newbies.

  • The Elements of Matlab Style (Johnson) is a great little book covering some tips for how to write the most elegant and readable Matlab code. Much of this generalizes to other programming languages. Read it after you have some familiarity with programming, since it assumes background knowledge.

  • Matlab for Behavioral Scientists (Rosenbaum). What it says. An introduction specifically tailored to introduce Matlab to social scientists.

Ticking off the boxes: Basic knowledge you should have

To go through the rest of the modules contained on this site, you should make sure you understand the following:

• Understand the most common data and variable types used in Matlab:

  • Scalars/numbers and numerical arrays

  • Character strings

  • Arrays and matrices as well as how to access their contents in full and in part

  • Cell arrays are like matrices, but can hold multiple different data types, including other cell arays.

  • Structures are like cell arays, but typically store and access data using fieldnames rather than numbers. This can be useful for grouping and working with data in different ways.

• Understand how commands are written and executed.

• Understand logical operations and their usage. This includes testing for equivalence (==, ~=), comparison (>, >=, <, <=), and multiple conditions (&, |). Learn about functions for comparing strings (e.g., isequal, strcmp, strcmpi)

• Understand how to control the order and flow of command execution

  • If/else statements

  • for loops

  • while loops

  • switch statements (used more rarely, but sometimes useful)

• Understand scripts and functions as well as the difference between them

• Understand the value and strategic use of comments (non-executable code that allows you to state, in English, what code is doing, to provide guides for more readable code)

• Understand file access and how to read and import different kinds of data

  • Know how to refer to the absolute or relative location of a file using path names. Understand how to navigate between different directories using the command cd

  • Work with text files

  • Work with images

• Understand how to use the debugging tools, using both the GUI and command-line functions like dbstop.

• Understand regular expressions and how to use them. Regular expressions can be incredibly helpful for finding full or partial text matches in data, and for replacing one set of text with another.

More advanced programming knowledge

None of the code presented in these pages requires the knowledge below. However, as you progress through coding skills, you may find these aspects of coding extremely helpful.

• If you are working with large datasets or computationally intensive analyses (e.g. machine learning, modeling, etc.), learning how to use Matlab's distributed computing capabilities is a must. Otherwise, you could be looking at your first grey hairs before an analysis runs to completion (assuming you don't already have them!).

  • Understand the uses of the parpool/parfor combo for breaking up jobs with a huge number of loop-iterations. For example, our lab uses this functionality to reduce the time taken to estimate complex computational models.

  • Understand how to run multiple instances of a function simultaneously and in parallel using the createJob/createTask combo. This can be useful if you've written a script/function that runs once per subject, but takes a long time to finish all the calculations for that subject. We often use functionality like this to process subjects through a neuroimaging pipeline in parallel, or to estimate parameters of a computational model on a per-subject basis.

return to main page