ans = 
Results for
I’ve installed Claude-code, MATLAB MCP Core Server, and now Puppeteer on my MacBook Pro. Puppeteer can navigate and operate web pages like Perplexity Comet or the new Claude Chrome Extension. The new wrinkle is MATLAB in the loop.
Claude-code and MATLAB MCP installation are described at Experiments with Claude code and MATLAB MPC Core Server . To install and configure Puppeteer, I used Claude App and its ability to use my MATLAB’s access to system files. The installation includes a Google Chrome for testing browser that is independent of (and does not interfere with) my normal Chrome browser. Puppeteer installation took just minutes of my approving various steps, and quitting and relaunchinbg Claude App. A minor hiccough was overwriting a special fetch connector configuration but that was readily fixed. The resulting linkage is Claude (cloud) ↔ Claude Desktop App ↔ MCP Server (local) ↔ Puppeteer ↔ Chrome for Testing (local) as well as the link to MATLAB on my laptop.
As a very first test, I selected Wikipedia from the Claude App suggestions. We navigated to a page Lorenz system in the Chrome for testing browser where Claude dismissed a prompt for donations to Wikipedia. ( I was like “What is Anthropic’s valuation and why didn’t you donate?” but said nothing.) and extracted content and summarized, many of the equations beautifully formated. I issued the following prompt: “Take a look at the differential equations there and at the example solutions and their parameter values. Then create a MATLAB script in my folder MATLAB/ClaudePuppeteer to reproduce the illustrations and run the script.” After clicking to approve various steps, presto.
The screen shot below shows 1) Claude App (upper left) after the process completed, 2) a MacOS Finder window showing the Lorentz Attractor.m in a folder ClaudePuppeteer that Claude had previously created for me to test Puppeteer functionality, 3) the Lorentz system wiki page in the Google Chrome for testing browser (lower right), 4) the LorentzAttractor script open in MATLAB (upper right), and 5) various figures created by the MATLAB Script.
Um, wow!
Introduction
MCP is an open protocol that can link Claude and other AI Apps to MATLAB using MATLAB MCP Core Server (released in Nov 2025). For an introduction, see Exploring the MATLAB Model Context Protocol (MCP) Core Server with Claude Desktop. Here, I describe my experience with installation and testing Claude-Code and MATLAB, a security concern, and in particular how I "taught" Claude to handle various MATLAB file formats.
Setup
A basic installation requires you download for your operating system claude-code, matlab-mcp-core-server, and node.js. One configuration is a terminal-launched claude connected to MATLAB. To connect Claude App to MATLAB requires an alternate configuration step and I recommend it for interative use. The configuration defines the default node/folder and MATLAB APP location.
I recommend using Claude itself to guide you through the installation and configuration steps for your operating system by providing terminal commands. I append Claude’s general description of installation for my APPLE Silicon laptop. Once set up, just ask in Claude App to do something in MATLAB and MATLAB App will be launched.
Security warning: Explore the following at your own risk.
When working with Claude App, Claude code, and MATLAB, you are granting Claude AI access to read and write files. By default, you must approve (one time or forever) any action so you hopefully don’t clobber files etc. Claude App believes it can not directly access file outside the top node defined in the setup. For this reason, I set the top node to be a folder ..../Documents/MATLAB. However, Claude inherits MATLAB App's command line privileges, typically your full system privileges. Claude can describe for you some work-arounds like a Docker container which might still be license validation compatible. I have not explored such options. During my setup, Claude just provided me terminal commands to copy and run. After setup, I've demonstrated it can run system level commands via matlab:evaluate_matlab_code and the MCP server. Be careful out there!
My first test
Claude can write a text-based .m script, execute it, collect text standard output from it, and open files it makes (or any file). It cannot access figures that you might see in MATLAB App unless they are saved as files or embedded in files. As we will see, the figures generated by a Live Script are saved in an Claude-accessible format when the Live Script is saved so the code need not itself export them.
In the screen shot below, the window at left is the Claude App after a successful connection. The MATLAB App window shows a script in the MATLAB editor that simulates a ballistics experiment, the script created successfully with a terminal-interfaced Claude and a simple prompt on the first try.
I deliberately but trivially broke this script using MATLAB App interactively by commenting out a needed variable g (acceleration of gravity) and saving the script to the edit was accessible to Claude. Using Claude App after its connection, I fixed the script with a simple prompt and ran it successfully to make the figure you see. The visible MATLAB didn’t know the code had been altered and fixed by Claude until I reloaded the file. Claude recommends plots be saved in PNG or JPEG, not PDF. It can describe in detail a plot in a PNG and thusly judge if the code is functioning correctly.
Live Scripts with Claude
What about Live Scripts (.mlx) and the (2025a) .m live? A .mlx file is a zipped package of files mixing code and images wtih XML markup. You can peek inside one and edit it directly without unzipping and rezipping it using a tool like BBEdit on a Mac, as shown below. This short test script has two interactive slider controls. You can in v2025+ now save a .mlx in a transportable .m Live text file format. The .mlx and .m Live formats have special markup for formatted text, interactive features like sliders, and figures.
Claude can convert a vanilla .m file to .mlx using matlab.internal.liveeditor.openAndSave(source.m, dest.mlx) and the reverse matlab.internal.liveeditor.openAndConvert('myfile.mlx', 'myfile.m’).
These functions do not support .m Live yet apparently. It would be great if they did.
Vanilla Claude doesn’t know all the details of the .mlx and .m formats yet. To teach Claude these formats, I created a simple .mlx with interactive features (sliders), saved it, and exported it to .m Live. I then sketched for Claude these structures. Claude digested these files and was then able to create such formats from scratch immediately. Wow! In the figure below, you see Claude on the left after creating a .mlx with sliders from scratch. The successfully executed Live Script appears in the MATLAB App.
Once Claude knows the .mlx format, it can peek inside a saved .mlx, process a PNG figure saved there, and describe it, as illustrated below. Here you see Claude describing a plot it found in the .mlx shown in the MATLAB window.
Wrap up
This summarizes my first look at using claude-code. The generation of simple MATLAB codes with the latest models is increasingly trivial with Claude, Perplexity, Open.ai GPT and other apps and the direct method of code and other file transfer between a native MATLAB and Claude seems to beat a copy-paste workflow. New methods to connect MATLAB Online with commercial AI's like Comet and Anthropic's Chrome extension seem to me more fluid and intuitive in that the user and the AI are co-working the same interface without copy-paste steps, but seem slower.
Appendix
Here are installation and setup instructions for Mac from Claude for reference.
Setting Up Claude Code + MATLAB MCP Server on Mac
Part 1: Claude Code
Install via npm:
bash
npm install -g @anthropic-ai/claude-code
Or via Homebrew:
bash
brew install claude-code
Run it:
bash
cd /path/to/your/project
claude
You'll need to authenticate with your Anthropic account on first run. Claude Code works in your terminal and can read/write files, run commands, and do agentic coding tasks.
Part 2: MATLAB MCP Server
Step 1: Download the server binary
Go to the MathWorks GitHub or File Exchange and download matlab-mcp-core-server for macOS (Apple Silicon or Intel depending on your Mac).
Step 2: Make it executable
bash
chmod +x ~/Downloads/matlab-mcp-core-server
Step 3: Create/edit Claude Desktop config
Edit ~/Library/Application Support/Claude/claude_desktop_config.json:
json
{
"mcpServers": {
"matlab": {
"command": "/Users/YOURUSERNAME/Downloads/matlab-mcp-core-server",
"args": ["--matlab-root", "/Applications/MATLAB_R2025a.app"]
},
"filesystem": {
"command": "npx",
"args": [
"@modelcontextprotocol/server-filesystem",
"/Users/YOURUSERNAME/Documents/MATLAB"
]
}
}
}
Replace YOURUSERNAME with your actual username, and adjust the MATLAB version if needed.
Step 4: Install Node.js (if not already)
bash
brew install node
Step 5: Restart Claude Desktop
Quit fully (Cmd+Q) and reopen. You should see a hammer/tools icon indicating MCP servers are connected.
Part 3: Verify Connection
In Claude Desktop, ask me to run MATLAB code. I should be able to execute:
matlab
disp('Hello from MATLAB!')
Troubleshooting
Check logs:
bash
cat ~/Library/Logs/Claude/mcp-server-matlab.log
cat ~/Library/Logs/Claude/mcp.log
Common issues:
- Missing --matlab-root argument → "no valid MATLAB environments found"
Connecting Claude App to MATLAB via MCP Server
Edit ~/Library/Application Support/Claude/claude_desktop_config.json:
json
{
"mcpServers": {
"filesystem": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-filesystem",
"/Users/YOURUSERNAME/Documents/MATLAB"
]
},
"matlab": {
"command": "/Users/YOURUSERNAME/Downloads/matlab-mcp-core-server",
"args": [
"--matlab-root", "/Applications/MATLAB_R2025a.app"
]
}
}
}
Then fully quit Claude Desktop (Cmd+Q) and reopen.
Comet browser can figure out and operate a user interface on the web including MATLAB Online. The screen shot shows MATLAB online to the left of the Comet AI. You see a test Live Script with sliders thjat Comet created in a folder (that it created). Comet is summarizing suggested improvements it requested of MATLAB Online's Copilot. Comet can plow into the arcane NASA astrophysical database interface SIMBAD, figure out how to grab information about, say, a star orbiting the black hole in the center of our galaxy and structure that information into a MATLAB data structure in a MATLAB script and run the script in MATLAB Online and display the results in the structure - it succeeded on the first try. It can do a Google Scholar citation tree search and park the results in MATLAB (success first try) or presumably in say MS Word in Office online. Presumably it can switch to a JuypterHub tab to then run a notebook or what you will. It can make mistakes, explore and discover GUI options, and recover e.g. use undo paste in MATLAB, operate in the command window, get help via the GUI if needed. Not at light speed but this seems enabling. Just released, Claude Chrome Extension Beta has similar capabilities.
Inspired in part by Christmas Trees, I'm curious about people's experience using AI to generate Matlab code.
1. Do you use AI to generate production code or just for experimentation/fun code?
2. Do you use the AI for a complete solution? Or is it more that the AI gets you most of the way there and you have to apply the finishing touches manually?
3. What level of quality would you consider the generated code? Does it follow "standard" Matlab coding practices? Is it well commented? Factored into modular functions? Argument checking? Memory efficient? Fast execution? Etc.?
4. Does the AI ever come up with a good or clever solution of which you wouldn't have thought or maybe of which you weren't even aware?
5. Is it easy/hard to express your requirements in a manner that the AI tool effectively translates into something useful?
6. Any other thoughts you'd care to share?
(Requested for newer MATLAB releases (e.g. R2026B), MATLAB Parallel Processing toolbox.)
Lower precision array types have been gaining more popularity over the years for deep learning. The current lowest precision built-in array type offered by MATLAB are 8-bit precision arrays, e.g. int8 and uint8. A good thing is that these 8-bit array types do have gpuArray support, meaning that one is able to design GPU MEX codes that take in these 8-bit arrays and reinterpret them bit-wise as other 8-bit array types, e.g. FP8, which is especially common array type used in modern day deep learning applications. I myself have used this to develop forward pass operations with 8-bit precision that are around twice as fast as 16-bit operations and with output arrays that still agree well with 16-bit outputs (measured with high cosine similarity). So the 8-bit support that MATLAB offers is already quite sufficient.
Recently, 4-bit precision array types have been shown also capable of being very useful in deep learning. These array types can be processed with Tensor Cores of more modern GPUs, such as NVIDIA's Blackwell architecture. However, MATLAB does not yet have a built-in 4-bit precision array type.
Just like MATLAB has int8 and uint8, both also with gpuArray support, it would also be nice to have MATLAB have int4 and uint4, also with gpuArray support.
I thought it would be fun to vibe code Christmas trees.
Model: Claude Sonnet4.5
Prompt: Write a MATLAB script that draws a Christmas tree. There should be decorations and presents too.
Result:
christmasTree()
function christmasTree()
% Christmas Tree with Decorations and Presents
% Creates a festive scene with a decorated tree and gift boxes
figure('Color', 'w', 'Position', [100, 100, 800, 900]);
hold on;
axis equal;
axis off;
xlim([-3, 3]);
ylim([0, 10]);
% Draw the tree (three layered triangular sections)
drawTreeSection(0, 2, 2.5, 2.5, [0.1, 0.5, 0.1]);
drawTreeSection(0, 3.5, 2, 2, [0.15, 0.55, 0.15]);
drawTreeSection(0, 4.8, 1.5, 1.8, [0.2, 0.6, 0.2]);
% Tree trunk
rectangle('Position', [-0.3, 0, 0.6, 2], 'FaceColor', [0.4, 0.2, 0.1], ...
'EdgeColor', 'none');
% Add decorations
addOrnaments();
addTinsel();
% Star on top
drawStar(0, 7.2, 0.4, [1, 0.84, 0]);
% Add presents under the tree
drawPresent(-1.8, 0, 0.6, 0.5, [0.8, 0.1, 0.1], [1, 0.84, 0]);
drawPresent(-0.8, 0, 0.5, 0.6, [0.1, 0.3, 0.8], [0.9, 0.9, 0.9]);
drawPresent(1.2, 0, 0.7, 0.4, [0.7, 0.5, 0.9], [0.2, 0.8, 0.2]);
drawPresent(0.2, 0, 0.4, 0.5, [1, 0.7, 0.1], [0.8, 0.1, 0.1]);
% Add some snow on the ground
scatter(randn(1, 100)*2.5, rand(1, 100)*0.3, 20, 'w', 'filled', ...
'MarkerEdgeColor', [0.8, 0.8, 0.8]);
title('Merry Christmas!', 'FontSize', 24, 'Color', [0.8, 0.1, 0.1], ...
'FontWeight', 'bold');
hold off;
end
function drawTreeSection(x, y, width, height, color)
% Draw a triangular section of the tree
xPoints = [x - width, x + width, x];
yPoints = [y, y, y + height];
fill(xPoints, yPoints, color, 'EdgeColor', [0.05, 0.3, 0.05], ...
'LineWidth', 1.5);
end
function addOrnaments()
% Add colorful baubles to the tree
ornamentColors = [
1, 0, 0; % Red
0, 0, 1; % Blue
1, 0.84, 0; % Gold
0.9, 0.1, 0.5;% Pink
0.5, 0, 0.5 % Purple
];
% Ornament positions (x, y coordinates)
positions = [
-1.5, 2.5; 1.5, 2.5; 0, 3.2;
-1.2, 3.8; 1.2, 3.8; -0.6, 4.2; 0.6, 4.2;
-0.8, 5.2; 0.8, 5.2; 0, 5.8; -0.4, 6.3; 0.4, 6.3
];
for i = 1:size(positions, 1)
colorIdx = mod(i-1, size(ornamentColors, 1)) + 1;
x = positions(i, 1);
y = positions(i, 2);
% Draw bauble
theta = linspace(0, 2*pi, 50);
radius = 0.15;
xCirc = x + radius * cos(theta);
yCirc = y + radius * sin(theta);
fill(xCirc, yCirc, ornamentColors(colorIdx, :), ...
'EdgeColor', ornamentColors(colorIdx, :)*0.7, 'LineWidth', 1.5);
% Add highlight for shine
highlightX = x - radius*0.4 + radius*0.3*cos(theta(1:15));
highlightY = y + radius*0.4 + radius*0.3*sin(theta(1:15));
fill(highlightX, highlightY, [1, 1, 1], 'EdgeColor', 'none', ...
'FaceAlpha', 0.6);
end
end
function addTinsel()
% Add sparkly tinsel effect
tinselX = [-1.8, -1.2, -0.5, 0.2, 0.9, 1.6];
tinselY = [2.8, 3.5, 4.0, 4.5, 5.0, 5.5];
for i = 1:length(tinselX)
plot(tinselX(i) + randn(1, 5)*0.1, tinselY(i) + rand(1, 5)*0.2, ...
'*', 'Color', [1, 0.84, 0], 'MarkerSize', 8, 'LineWidth', 1.5);
end
end
function drawStar(x, y, size, color)
% Draw a 5-pointed star on top of the tree
angles = linspace(0, 2*pi, 11);
radii = repmat([size, size*0.4], 1, 5);
radii = [radii, radii(1)];
xStar = x + radii .* cos(angles - pi/2);
yStar = y + radii .* sin(angles - pi/2);
fill(xStar, yStar, color, 'EdgeColor', [0.8, 0.6, 0], 'LineWidth', 2);
% Add sparkle
plot(x, y, '*', 'Color', [1, 1, 1], 'MarkerSize', 15, 'LineWidth', 2);
end
function drawPresent(x, y, width, height, boxColor, ribbonColor)
% Draw a gift box with ribbon and bow
% Box
rectangle('Position', [x, y, width, height], 'FaceColor', boxColor, ...
'EdgeColor', boxColor*0.7, 'LineWidth', 2);
% Vertical ribbon
ribbonWidth = width * 0.15;
rectangle('Position', [x + width/2 - ribbonWidth/2, y, ribbonWidth, height], ...
'FaceColor', ribbonColor, 'EdgeColor', 'none');
% Horizontal ribbon
ribbonHeight = height * 0.15;
rectangle('Position', [x, y + height/2 - ribbonHeight/2, width, ribbonHeight], ...
'FaceColor', ribbonColor, 'EdgeColor', 'none');
% Bow on top
bowX = x + width/2;
bowY = y + height;
bowSize = width * 0.2;
% Left loop
theta = linspace(0, pi, 30);
fill(bowX - bowSize*0.3 + bowSize*0.5*cos(theta), ...
bowY + bowSize*0.5*sin(theta), ribbonColor, 'EdgeColor', 'none');
% Right loop
fill(bowX + bowSize*0.3 + bowSize*0.5*cos(theta), ...
bowY + bowSize*0.5*sin(theta), ribbonColor, 'EdgeColor', 'none');
% Center knot
theta = linspace(0, 2*pi, 30);
fill(bowX + bowSize*0.25*cos(theta), bowY + bowSize*0.25*sin(theta), ...
ribbonColor*0.8, 'EdgeColor', 'none');
end
I like this quote, what do you think?
"If the part of programming you enjoy most is the physical act of writing code, then agents will feel beside the point. You’re already where you want to be, even just with some Copilot or Cursor-style intelligent code auto completion, which makes you faster while still leaving you fully in the driver’s seat about the code that gets written.
But if the part you care about is the decision-making around the code, agents feel like they clear space. They take care of the mechanical expression and leave you with judgment, tradeoffs, and intent. Because truly, for someone at my experience level, that is my core value offering anyway. When I spend time actually typing code these days with my own fingers, it feels like a waste of my time."
— Obie Fernandez, What happens when the coding becomes the least interesting part of the work
I can't believe someone put time into this ;-)
In a recent blog post, @Guy Rouleau writes about the new Simulink Copilot Beta. Sign ups are on the Copilot Beta page below. Let him know what you think.
Guy's Blog Post - https://blogs.mathworks.com/simulink/2025/12/01/a-copilot-for-simulink/
Simulink Copilot Beta - https://www.mathworks.com/products/simulink-copilot.html
The formula comes from @yuruyurau. (https://x.com/yuruyurau)
digital life 1
figure('Position',[300,50,900,900], 'Color','k');
axes(gcf, 'NextPlot','add', 'Position',[0,0,1,1], 'Color','k');
axis([0, 400, 0, 400])
SHdl = scatter([], [], 2, 'filled','o','w', 'MarkerEdgeColor','none', 'MarkerFaceAlpha',.4);
t = 0;
i = 0:2e4;
x = mod(i, 100);
y = floor(i./100);
k = x./4 - 12.5;
e = y./9 + 5;
o = vecnorm([k; e])./9;
while true
t = t + pi/90;
q = x + 99 + tan(1./k) + o.*k.*(cos(e.*9)./4 + cos(y./2)).*sin(o.*4 - t);
c = o.*e./30 - t./8;
SHdl.XData = (q.*0.7.*sin(c)) + 9.*cos(y./19 + t) + 200;
SHdl.YData = 200 + (q./2.*cos(c));
drawnow
end
digital life 2
figure('Position',[300,50,900,900], 'Color','k');
axes(gcf, 'NextPlot','add', 'Position',[0,0,1,1], 'Color','k');
axis([0, 400, 0, 400])
SHdl = scatter([], [], 2, 'filled','o','w', 'MarkerEdgeColor','none', 'MarkerFaceAlpha',.4);
t = 0;
i = 0:1e4;
x = i;
y = i./235;
e = y./8 - 13;
while true
t = t + pi/240;
k = (4 + sin(y.*2 - t).*3).*cos(x./29);
d = vecnorm([k; e]);
q = 3.*sin(k.*2) + 0.3./k + sin(y./25).*k.*(9 + 4.*sin(e.*9 - d.*3 + t.*2));
SHdl.XData = q + 30.*cos(d - t) + 200;
SHdl.YData = 620 - q.*sin(d - t) - d.*39;
drawnow
end
digital life 3
figure('Position',[300,50,900,900], 'Color','k');
axes(gcf, 'NextPlot','add', 'Position',[0,0,1,1], 'Color','k');
axis([0, 400, 0, 400])
SHdl = scatter([], [], 1, 'filled','o','w', 'MarkerEdgeColor','none', 'MarkerFaceAlpha',.4);
t = 0;
i = 0:1e4;
x = mod(i, 200);
y = i./43;
k = 5.*cos(x./14).*cos(y./30);
e = y./8 - 13;
d = (k.^2 + e.^2)./59 + 4;
a = atan2(k, e);
while true
t = t + pi/20;
q = 60 - 3.*sin(a.*e) + k.*(3 + 4./d.*sin(d.^2 - t.*2));
c = d./2 + e./99 - t./18;
SHdl.XData = q.*sin(c) + 200;
SHdl.YData = (q + d.*9).*cos(c) + 200;
drawnow; pause(1e-2)
end
digital life 4
figure('Position',[300,50,900,900], 'Color','k');
axes(gcf, 'NextPlot','add', 'Position',[0,0,1,1], 'Color','k');
axis([0, 400, 0, 400])
SHdl = scatter([], [], 1, 'filled','o','w', 'MarkerEdgeColor','none', 'MarkerFaceAlpha',.4);
t = 0;
i = 0:4e4;
x = mod(i, 200);
y = i./200;
k = x./8 - 12.5;
e = y./8 - 12.5;
o = (k.^2 + e.^2)./169;
d = .5 + 5.*cos(o);
while true
t = t + pi/120;
SHdl.XData = x + d.*k.*sin(d.*2 + o + t) + e.*cos(e + t) + 100;
SHdl.YData = y./4 - o.*135 + d.*6.*cos(d.*3 + o.*9 + t) + 275;
SHdl.CData = ((d.*sin(k).*sin(t.*4 + e)).^2).'.*[1,1,1];
drawnow;
end
digital life 5
figure('Position',[300,50,900,900], 'Color','k');
axes(gcf, 'NextPlot','add', 'Position',[0,0,1,1], 'Color','k');
axis([0, 400, 0, 400])
SHdl = scatter([], [], 1, 'filled','o','w',...
'MarkerEdgeColor','none', 'MarkerFaceAlpha',.4);
t = 0;
i = 0:1e4;
x = mod(i, 200);
y = i./55;
k = 9.*cos(x./8);
e = y./8 - 12.5;
while true
t = t + pi/120;
d = (k.^2 + e.^2)./99 + sin(t)./6 + .5;
q = 99 - e.*sin(atan2(k, e).*7)./d + k.*(3 + cos(d.^2 - t).*2);
c = d./2 + e./69 - t./16;
SHdl.XData = q.*sin(c) + 200;
SHdl.YData = (q + 19.*d).*cos(c) + 200;
drawnow;
end
digital life 6
clc; clear
figure('Position',[300,50,900,900], 'Color','k');
axes(gcf, 'NextPlot','add', 'Position',[0,0,1,1], 'Color','k');
axis([0, 400, 0, 400])
SHdl = scatter([], [], 2, 'filled','o','w', 'MarkerEdgeColor','none', 'MarkerFaceAlpha',.4);
t = 0;
i = 1:1e4;
y = i./790;
k = y; idx = y < 5;
k(idx) = 6 + sin(bitxor(floor(y(idx)), 1)).*6;
k(~idx) = 4 + cos(y(~idx));
while true
t = t + pi/90;
d = sqrt((k.*cos(i + t./4)).^2 + (y/3-13).^2);
q = y.*k.*cos(i + t./4)./5.*(2 + sin(d.*2 + y - t.*4));
c = d./3 - t./2 + mod(i, 2);
SHdl.XData = q + 90.*cos(c) + 200;
SHdl.YData = 400 - (q.*sin(c) + d.*29 - 170);
drawnow; pause(1e-2)
end
digital life 7
clc; clear
figure('Position',[300,50,900,900], 'Color','k');
axes(gcf, 'NextPlot','add', 'Position',[0,0,1,1], 'Color','k');
axis([0, 400, 0, 400])
SHdl = scatter([], [], 2, 'filled','o','w', 'MarkerEdgeColor','none', 'MarkerFaceAlpha',.4);
t = 0;
i = 1:1e4;
y = i./345;
x = y; idx = y < 11;
x(idx) = 6 + sin(bitxor(floor(x(idx)), 8))*6;
x(~idx) = x(~idx)./5 + cos(x(~idx)./2);
e = y./7 - 13;
while true
t = t + pi/120;
k = x.*cos(i - t./4);
d = sqrt(k.^2 + e.^2) + sin(e./4 + t)./2;
q = y.*k./d.*(3 + sin(d.*2 + y./2 - t.*4));
c = d./2 + 1 - t./2;
SHdl.XData = q + 60.*cos(c) + 200;
SHdl.YData = 400 - (q.*sin(c) + d.*29 - 170);
drawnow; pause(5e-3)
end
digital life 8
clc; clear
figure('Position',[300,50,900,900], 'Color','k');
axes(gcf, 'NextPlot','add', 'Position',[0,0,1,1], 'Color','k');
axis([0, 400, 0, 400])
SHdl{6} = [];
for j = 1:6
SHdl{j} = scatter([], [], 2, 'filled','o','w', 'MarkerEdgeColor','none', 'MarkerFaceAlpha',.3);
end
t = 0;
i = 1:2e4;
k = mod(i, 25) - 12;
e = i./800; m = 200;
theta = pi/3;
R = [cos(theta) -sin(theta); sin(theta) cos(theta)];
while true
t = t + pi/240;
d = 7.*cos(sqrt(k.^2 + e.^2)./3 + t./2);
XY = [k.*4 + d.*k.*sin(d + e./9 + t);
e.*2 - d.*9 - d.*9.*cos(d + t)];
for j = 1:6
XY = R*XY;
SHdl{j}.XData = XY(1,:) + m;
SHdl{j}.YData = XY(2,:) + m;
end
drawnow;
end
digital life 9
clc; clear
figure('Position',[300,50,900,900], 'Color','k');
axes(gcf, 'NextPlot','add', 'Position',[0,0,1,1], 'Color','k');
axis([0, 400, 0, 400])
SHdl{14} = [];
for j = 1:14
SHdl{j} = scatter([], [], 2, 'filled','o','w', 'MarkerEdgeColor','none', 'MarkerFaceAlpha',.1);
end
t = 0;
i = 1:2e4;
k = mod(i, 50) - 25;
e = i./1100; m = 200;
theta = pi/7;
R = [cos(theta) -sin(theta); sin(theta) cos(theta)];
while true
t = t + pi/240;
d = 5.*cos(sqrt(k.^2 + e.^2) - t + mod(i, 2));
XY = [k + k.*d./6.*sin(d + e./3 + t);
90 + e.*d - e./d.*2.*cos(d + t)];
for j = 1:14
XY = R*XY;
SHdl{j}.XData = XY(1,:) + m;
SHdl{j}.YData = XY(2,:) + m;
end
drawnow;
end
% Recreation of Saturn photo
figure('Color', 'k', 'Position', [100, 100, 800, 800]);
ax = axes('Color', 'k', 'XColor', 'none', 'YColor', 'none', 'ZColor', 'none');
hold on;
% Create the planet sphere
[x, y, z] = sphere(150);
% Saturn colors - pale yellow/cream gradient
saturn_radius = 1;
% Create color data based on latitude for gradient effect
lat = asin(z);
color_data = rescale(lat, 0.3, 0.9);
% Plot Saturn with smooth shading
planet = surf(x*saturn_radius, y*saturn_radius, z*saturn_radius, ...
color_data, ...
'EdgeColor', 'none', ...
'FaceColor', 'interp', ...
'FaceLighting', 'gouraud', ...
'AmbientStrength', 0.3, ...
'DiffuseStrength', 0.6, ...
'SpecularStrength', 0.1);
% Use a cream/pale yellow colormap for Saturn
cream_map = [linspace(0.4, 0.95, 256)', ...
linspace(0.35, 0.9, 256)', ...
linspace(0.2, 0.7, 256)'];
colormap(cream_map);
% Create the ring system
n_points = 300;
theta = linspace(0, 2*pi, n_points);
% Define ring structure (inner radius, outer radius, brightness)
rings = [
1.2, 1.4, 0.7; % Inner ring
1.45, 1.65, 0.8; % A ring
1.7, 1.85, 0.5; % Cassini division (darker)
1.9, 2.3, 0.9; % B ring (brightest)
2.35, 2.5, 0.6; % C ring
2.55, 2.8, 0.4; % Outer rings (fainter)
];
% Create rings as patches
for i = 1:size(rings, 1)
r_inner = rings(i, 1);
r_outer = rings(i, 2);
brightness = rings(i, 3);
% Create ring coordinates
x_inner = r_inner * cos(theta);
y_inner = r_inner * sin(theta);
x_outer = r_outer * cos(theta);
y_outer = r_outer * sin(theta);
% Front side of rings
ring_x = [x_inner, fliplr(x_outer)];
ring_y = [y_inner, fliplr(y_outer)];
ring_z = zeros(size(ring_x));
% Color based on brightness
ring_color = brightness * [0.9, 0.85, 0.7];
fill3(ring_x, ring_y, ring_z, ring_color, ...
'EdgeColor', 'none', ...
'FaceAlpha', 0.7, ...
'FaceLighting', 'gouraud', ...
'AmbientStrength', 0.5);
end
% Add some texture/gaps in the rings using scatter
n_particles = 3000;
r_particles = 1.2 + rand(1, n_particles) * 1.6;
theta_particles = rand(1, n_particles) * 2 * pi;
x_particles = r_particles .* cos(theta_particles);
y_particles = r_particles .* sin(theta_particles);
z_particles = (rand(1, n_particles) - 0.5) * 0.02;
% Vary particle brightness
particle_colors = repmat([0.8, 0.75, 0.6], n_particles, 1) .* ...
(0.5 + 0.5*rand(n_particles, 1));
scatter3(x_particles, y_particles, z_particles, 1, particle_colors, ...
'filled', 'MarkerFaceAlpha', 0.3);
% Add dramatic outer halo effect - multiple layers extending far out
n_glow = 20;
for i = 1:n_glow
glow_radius = 1 + i*0.35; % Extend much farther
alpha_val = 0.08 / sqrt(i); % More visible, slower falloff
% Color gradient from cream to blue/purple at outer edges
if i <= 8
glow_color = [0.9, 0.85, 0.7]; % Warm cream/yellow
else
% Gradually shift to cooler colors
mix = (i - 8) / (n_glow - 8);
glow_color = (1-mix)*[0.9, 0.85, 0.7] + mix*[0.6, 0.65, 0.85];
end
surf(x*glow_radius, y*glow_radius, z*glow_radius, ...
ones(size(x)), ...
'EdgeColor', 'none', ...
'FaceColor', glow_color, ...
'FaceAlpha', alpha_val, ...
'FaceLighting', 'none');
end
% Add extensive glow to rings - make it much more dramatic
n_ring_glow = 12;
for i = 1:n_ring_glow
glow_scale = 1 + i*0.15; % Extend farther
alpha_ring = 0.12 / sqrt(i); % More visible
for j = 1:size(rings, 1)
r_inner = rings(j, 1) * glow_scale;
r_outer = rings(j, 2) * glow_scale;
brightness = rings(j, 3) * 0.5 / sqrt(i);
x_inner = r_inner * cos(theta);
y_inner = r_inner * sin(theta);
x_outer = r_outer * cos(theta);
y_outer = r_outer * sin(theta);
ring_x = [x_inner, fliplr(x_outer)];
ring_y = [y_inner, fliplr(y_outer)];
ring_z = zeros(size(ring_x));
% Color gradient for ring glow
if i <= 6
ring_color = brightness * [0.9, 0.85, 0.7];
else
mix = (i - 6) / (n_ring_glow - 6);
ring_color = brightness * ((1-mix)*[0.9, 0.85, 0.7] + mix*[0.65, 0.7, 0.9]);
end
fill3(ring_x, ring_y, ring_z, ring_color, ...
'EdgeColor', 'none', ...
'FaceAlpha', alpha_ring, ...
'FaceLighting', 'none');
end
end
% Add diffuse glow particles for atmospheric effect
n_glow_particles = 8000;
glow_radius_particles = 1.5 + rand(1, n_glow_particles) * 5;
theta_glow = rand(1, n_glow_particles) * 2 * pi;
phi_glow = acos(2*rand(1, n_glow_particles) - 1);
x_glow = glow_radius_particles .* sin(phi_glow) .* cos(theta_glow);
y_glow = glow_radius_particles .* sin(phi_glow) .* sin(theta_glow);
z_glow = glow_radius_particles .* cos(phi_glow);
% Color particles based on distance - cooler colors farther out
particle_glow_colors = zeros(n_glow_particles, 3);
for i = 1:n_glow_particles
dist = glow_radius_particles(i);
if dist < 3
particle_glow_colors(i,:) = [0.9, 0.85, 0.7];
else
mix = (dist - 3) / 4;
particle_glow_colors(i,:) = (1-mix)*[0.9, 0.85, 0.7] + mix*[0.5, 0.6, 0.9];
end
end
scatter3(x_glow, y_glow, z_glow, rand(1, n_glow_particles)*2+0.5, ...
particle_glow_colors, 'filled', 'MarkerFaceAlpha', 0.05);
% Lighting setup
light('Position', [-3, -2, 4], 'Style', 'infinite', ...
'Color', [1, 1, 0.95]);
light('Position', [2, 3, 2], 'Style', 'infinite', ...
'Color', [0.3, 0.3, 0.4]);
% Camera and view settings
axis equal off;
view([-35, 25]); % Angle to match saturn_photo.jpg - more dramatic tilt
camva(10); % Field of view - slightly wider to show full halo
xlim([-8, 8]); % Expanded to show outer halo
ylim([-8, 8]);
zlim([-8, 8]);
% Material properties
material dull;
title('Saturn - Left click: Rotate | Right click: Pan | Scroll: Zoom', 'Color', 'w', 'FontSize', 12);
% Enable interactive camera controls
cameratoolbar('Show');
cameratoolbar('SetMode', 'orbit'); % Start in rotation mode
% Custom mouse controls
set(gcf, 'WindowButtonDownFcn', @mouseDown);
function mouseDown(src, ~)
selType = get(src, 'SelectionType');
switch selType
case 'normal' % Left click - rotate
cameratoolbar('SetMode', 'orbit');
rotate3d on;
case 'alt' % Right click - pan
cameratoolbar('SetMode', 'pan');
pan on;
end
end
Experimenting with Agentic AI
44%
I am an AI skeptic
0%
AI is banned at work
11%
I am happy with Conversational AI
44%
9 votes
Pure Matlab
82%
Simulink
18%
11 votes
Jorge Bernal-AlvizJorge Bernal-Alviz shared the following code that requires R2025a or later:
Test()
function Test()
duration = 10;
numFrames = 800;
frameInterval = duration / numFrames;
w = 400;
t = 0;
i_vals = 1:10000;
x_vals = i_vals;
y_vals = i_vals / 235;
r = linspace(0, 1, 300)';
g = linspace(0, 0.1, 300)';
b = linspace(1, 0, 300)';
r = r * 0.8 + 0.1;
g = g * 0.6 + 0.1;
b = b * 0.9 + 0.1;
customColormap = [r, g, b];
figure('Position', [100, 100, w, w], 'Color', [0, 0, 0]);
axis equal;
axis off;
xlim([0, w]);
ylim([0, w]);
hold on;
colormap default;
colormap(customColormap);
plothandle = scatter([], [], 1, 'filled', 'MarkerFaceAlpha', 0.12);
for i = 1:numFrames
t = t + pi/240;
k = (4 + 3 * sin(y_vals * 2 - t)) .* cos(x_vals / 29);
e = y_vals / 8 - 13;
d = sqrt(k.^2 + e.^2);
c = d - t;
q = 3 * sin(2 * k) + 0.3 ./ (k + 1e-10) + ...
sin(y_vals / 25) .* k .* (9 + 4 * sin(9 * e - 3 * d + 2 * t));
points_x = q + 30 * cos(c) + 200;
points_y = q .* sin(c) + 39 * d - 220;
points_y = w - points_y;
CData = (1 + sin(0.1 * (d - t))) / 3;
CData = max(0, min(1, CData));
set(plothandle, 'XData', points_x, 'YData', points_y, 'CData', CData);
brightness = 0.5 + 0.3 * sin(t * 0.2);
set(plothandle, 'MarkerFaceAlpha', brightness);
drawnow;
pause(frameInterval);
end
end
From my experience, MATLAB's Deep Learning Toolbox is quite user-friendly, but it still falls short of libraries like PyTorch in many respects. Most users tend to choose PyTorch because of its flexibility, efficiency, and rich support for many mathematical operators. In recent years, the number of dlarray-compatible mathematical functions added to the toolbox has been very limited, which makes it difficult to experiment with many custom networks. For example, svd is currently not supported for dlarray inputs.
This link (List of Functions with dlarray Support - MATLAB & Simulink) lists all functions that support dlarray as of R2026a — only around 200 functions (including toolbox-specific ones). I would like to see support for many more fundamental mathematical functions so that users have greater freedom when building and researching custom models. For context, the core MATLAB mathematics module contains roughly 600 functions, and many application domains build on that foundation.
I hope MathWorks will prioritize and accelerate expanding dlarray support for basic math functions. Doing so would significantly increase the Deep Learning Toolbox's utility and appeal for researchers and practitioners.
Thank you.
Run MATLAB using AI applications by leveraging MCP. This MCP server for MATLAB supports a wide range of coding agents like Claude Code and Visual Studio Code.
Check it out and share your experiences below. Have fun!
GitHub repo: https://github.com/matlab/matlab-mcp-core-server
Yann Debray's blog post: https://blogs.mathworks.com/deep-learning/2025/11/03/releasing-the-matlab-mcp-core-server-on-github/
I'm working on training neural networks without backpropagation / automatic differentiation, using locally derived analytic forms of update rules. Given that this allows a direct formula to be derived for the update rule, it removes alot of the overhead that is otherwise required from automatic differentiation.
However, matlab's functionalities for neural networks are currently solely based around backpropagation and automatic differentiation, such as the dlgradient function and requiring everything to be dlarrays during training.
I have two main requests, specifically for functions that perform a single operation within a single layer of a neural network, such as "dlconv", "fullyconnect", "maxpool", "avgpool", "relu", etc:
- these functions should also allow normal gpuArray data instead of requiring everything to be dlarrays.
- these functions are currently designed to only perform the forward pass. I request that these also be designed to perform the backward pass if user requests. There can be another input user flag that can be "forward" (default) or "backward", and then the function should have all the necessary inputs to perform that operation (e.g. for "avgpool" forward pass it only needs the avgpool input data and the avgpool parameters, but for the "avgpool" backward pass it needs the deriviative w.r.t. the avgpool output data, the avgpool parameters, and the original data dimensions). I know that there is a maxunpool function that achieves this for maxpool, but it has significant issues when trying to use it this way instead of by backpropagation in a dlgradient type layer, see (https://www.mathworks.com/matlabcentral/answers/2179587-making-a-custom-way-to-train-cnns-and-i-am-noticing-that-avgpool-is-significantly-faster-than-maxpo?s_tid=srchtitle).
I don't know how many people would benefit from this feature, and someone could always spend their time creating these functionalities themselves by matlab scripts, cuDNN mex, etc., but regardless it would be nice for matlab to have this allowable for more customizable neural net training.
Inspired by @xingxingcui's post about old MATLAB versions and @유장's post about an old Easter egg, I thought it might be fun to share some MATLAB-Old-Timer Stories™.
Back in the early 90s, MATLAB had been ported to MacOS, but there were some interesting wrinkles. One that kept me earning my money as a computer lab tutor was that MATLAB required file names to follow Windows standards - no spaces or other special characters. But on a Mac, nothing stopped you from naming your script "hello world - 123.m". The problem came when you tried to run it. MATLAB was essentially doing an eval on the script name, assuming the file name would follow Windows (and MATLAB) naming rules.
So now imagine a lab full of students taking a university course. As is common in many universities, the course was given a numeric code. For whatever historical reason, my school at that time was also using numeric codes for the departments. Despite being told the rules for naming scripts, many students would default to something like "26.165 - 1.1" for problem one on HW1 for the intro applied math course 26.165.
No matter what they did in their script, when they ran it, MATLAB would just say "ans = 25.0650".
Nothing brings you more MATLAB-god credibility as a student tutor than walking over to someone's computer, taking one look at their output, saying "rename your file", and walking away like a boss.
It was 2010 when I was a sophomore in university. I chose to learn MATLAB because of a mathematical modeling competition, and the university provided MATLAB 7.0, a very classic release. To get started, I borrowed many MATLAB books from the library and began by learning simple numerical calculations, plotting, and solving equations. Gradually I was drawn in by MATLAB’s powerful capabilities and became interested; I often used it as a big calculator for fun. That version didn’t have MATLAB Live Script; instead it used MATLAB Notebook (M-Book), which allowed MATLAB functions to be used directly within Microsoft Word, and it also had the Symbolic Math Toolbox’s MuPAD interactive environment. These were later gradually replaced by Live Scripts introduced in R2016a. There are many similar examples...
Out of curiosity, I still have screenshots on my computer showing MATLAB 7.0 running compatibly. I’d love to hear your thoughts?
Edit 15 Oct 2025: Removed incorrect code. Replaced symmatrix2sym and symfunmatrix2symfun with sym and symfun respectively (latter supported as of 2024b).
The Symbolic Math Toolbox does not have its own dot and and cross functions. That's o.k. (maybe) for garden variety vectors of sym objects where those operations get shipped off to the base Matlab functions
x = sym('x',[3,1]); y = sym('y',[3,1]);
which dot(x,y)
dot(x,y)
which cross(x,y)
cross(x,y)
clearvars
x = symmatrix('x',[3,1]); y = symmatrix('y',[3,1]);
z = symmatrix('z',[1,1]);
sympref('AbbreviateOutput',false);
dot() expands the result, which isn't really desirable for exposition.
eqn = z == dot(x,y)
Also, dot() returns the the result in terms of the conjugate of x, which can't be simplifed away at the symmatrix level
assumeAlso(sym(x),'real')
class(eqn)
try
eqn = z == simplify(dot(x,y))
catch ME
ME.message
end
To get rid of the conjugate, we have to resort to sym
eqn = simplify(sym(eqn))
but again we are in expanded form, which defeats the purpose of symmatrix (et. al.)
But at least we can do this to get a nice equation
eqn = z == x.'*y
dot errors with symfunmatrix inputs
clearvars
syms t real
x = symfunmatrix('x(t)',t,[3,1]); y = symfunmatrix('y(t)',t,[3,1]);
try
dot(x,y)
catch ME
ME.message
end
Cross works (accidentally IMO) with symmatrix, but expands the result, which isn't really desirable for exposition
clearvars
x = symmatrix('x',[3,1]); y = symmatrix('y',[3,1]);
z = symmatrix('z',[3,1]);
eqn = z == cross(x,y)
And it doesn't work at all if an input is a symfunmatrix
syms t
w = symfunmatrix('w(t)',t,[3,1]);
try
eqn = z == cross(x,w);
catch ME
ME.message
end
In the latter case we can expand with
eqn = z == cross(sym(x),symfun(w)) % x has to be converted
But we can't do the same with dot (as shown above, dot doesn't like symfun inputs)
try
eqn = z == dot(sym(x),symfun(w))
catch ME
ME.message
end
Looks like the only choice for dot with symfunmatrix is to write it by hand at the matrix level
x.'*w
or at the sym/symfun level
sym(x).'*symfun(w) % assuming x is real
Ideally, I'd like to see dot and cross implemented for symmatrix and symfunmatrix types where neither function would evaluate, i.e., expand, until both arguments are subs-ed with sym or symfun objects of appropriate dimension.
Also, it would be nice if symmatrix could be assumed to be real. Is there a reason why being able to do so wouldn't make sense?
try
assume(x,'real')
catch ME
ME.message
end
