The ACM Special Interest Group on Programming Languages, SIGPLAN, expects to hold the fourth in a series of conferences on the History of Programming Languages in 2020, see HOPL-IV. The first drafts of papers are to be submitted by August 2018. That long lead time gives me the opportunity to write a detailed history of MATLAB. I plan to write the paper in sections, which I'll post in this blog as they are available. This is the seventh, and final, installment.

MATLAB has evolved over the past 35 years into a rich technical computing environment. Here is an overview of some of the tools available.

Contents

Desktop

The MATLAB desktop was introduced in 2000. Here is the default layout of the desktop. This snapshot was taken while I was working on my previous post. Four panels are visible, the current folder viewer on the left, the workspace viewer on the right, the editor/debugger in the top center, and the traditional command window in the bottom center. A file viewer and a command history window are not in the default layout, but can be included in personalized layouts.

Any of the panels can be closed, or undocked into a standalone window. I have two screens available on the desk in my office. I usually put the command window on one screen and the editor on the other.

Workspace

Here are the contents of the workspace when I execute the source script for my previous post. If I click on any of the variables, it will be opened in an appropriate viewer

Editor/Debugger

When I work on this blog, I spend most of my time using this MATLAB editor. The source for each post is an executable script, with descriptive text in comments and section headings in comments beginning with double percent signs, %%. The final production of a post is initiated with the "Publish" tab. I have pretty much abandoned my old Unix text editor, "vi", after many years of faithful service.

Live Editor

The Live Editor was introduced in 2016 and is still under active development. MATLAB input, output and graphics are combined in a single interactive document. The document may be exported to HTML, PDF, or LaTeX.

Here are a few examples. The table generated by the small gradebook from my previous post is nicely formatted.

Expressions from the Symbolic Toolbox are typeset.

And graphical output may be included in the document.

Parallel computing

The Parallel Computing Toolbox (PCT) was introduced at the Super Computing conference in 2004. The next year, at SC05, Bill Gates gave the keynote talk, using MATLAB to demonstrate Microsoft's entry into High Performance Computing.

The toolbox supports coarse grained, distributed memory parallelism by running many MATLAB workers on many machines in a cluster, or on many cores in a single machine. Originally MATLAB used MPI for the message passing, but in more recent versions of the toolbox our own utilities have replaced MPI.

By far the most popular feature of the PCT is the parallel for loop command, parfor.

GPUs

Support for Graphics Processing Units was added to the Parallel Computing Toolbox in 2010. Eight years later, in release R2018a, the gpuarray has grown to have 385 associated methods, including all of the familiar matrix computations, lu, eig, svd and mldivide (backslash).

Toolboxes

Much of modern MATLAB's power derives from the toolboxes available for specialized applications. In release 2018a there are 63 of them. Here is the list.

Parallel Computing

Math, Statistics, and Optimization

Control Systems

Signal Processing and Wireless Communications

Image Processing and Computer Vision

Test and Measurement

Computational Finance

Computational Biology

Code Generation

Application Deployment

Database Access and Reporting

Get the MATLAB code

Published with MATLAB® R2018a