🚀 Clustering made interactive in Scilab

General usage
1

Added in Scilab 2026.1.0, this demo combines the best of both worlds: HTML for a sleek, modern UI — and Scilab for the numerical heavy lifting. Every interaction in the HTML panels triggers computations in Scilab’s engine, with results plotted live.

Three popular clustering algorithms run side by side on a single dataset:
:blue_circle: DBSCAN — density-based, detects noise and arbitrary shapes
:green_circle: Mean Shift — finds clusters without specifying their number
:orange_circle: K-Means — the classic partitioning approach

Adjust the parameters, regenerate the data, and instantly see how each method responds — a hands-on way to build intuition about what each algorithm does, and where it shines.

:backhand_index_pointing_right: Try it out and explore your own data!

5 Likes
2

i have a problem with kmeans

Operator -: Wrong dimensions for operation [5x2] - [6x2], same dimensions expected.

[idx,centroid] = kmeans(K, m);

m=6 and
K as CSV

K.csv (17.1 KB)

first time error warnin, second time it works, seem some old configuration remained stored in the function.

3

Hello Pascal,

I reproduce your issue. I have fixed it here

Adeline

1 Like
4

Thank you, for your time :slight_smile:

1 Like