Scilab — The Complete Guide: Free Numerical Computing and Simulation With Xcos
Scilab is an open-source numerical computing environment, long established and used in universities and industry, mainly in Europe. It is similar to MATLAB in concept — matrices, functions, plots — but its standout advantage is Xcos: a block-diagram simulation environment of the kind familiar from Simulink.
Scilab
Dassault Systèmes / Scilab Enterprises
Numerical computing with Xcos — block-diagram simulation of dynamic systems alongside scripting.
Get it from the vendorWhat is in Scilab
- Numerical computing — linear algebra, differential equations, optimisation, interpolation and statistics.
- Plots — 2D and 3D, with export to image and PDF.
- Xcos — building a system from blocks: integrators, transfer functions, PID controllers, sources and scopes — and running a time simulation.
- Control and signal processing — tools for frequency response, stability, filters and FFT.
- ATOMS — a package manager for additional extensions.
Scilab versus MATLAB and Octave
Worth knowing: Scilab's syntax is not compatible with MATLAB. An .m file will not run as is — there is a conversion tool, but fixes are usually needed. If you have a lot of existing MATLAB code, GNU Octave fits better. If you need block-based simulation, Scilab with Xcos is the leading free alternative to Simulink.
GNU Octave
GNU Project
Numerical computing in MATLAB-compatible syntax — matrices, equations, signals and plots.
Get it from the vendorWhat's missing
The Xcos block library is smaller than Simulink's, and most of MATLAB's advanced toolboxes have no equivalent. The community is smaller than Python's or Octave's, so there are fewer examples online.
Who it suits
Students and engineers in control, electrical and mechanical engineering who need dynamic system simulation, lecturers in control and systems courses, and anyone looking for a Simulink alternative without a licence.
Alternatives and companion tools
For existing MATLAB code — GNU Octave. For equation-based multi-domain physical models — OpenModelica. For the widest ecosystem — Python. And for symbolic maths — wxMaxima.
OpenModelica
Open Source Modelica Consortium
Simulate multi-domain physical systems with the Modelica language.
Get it from the vendor
Python
Python Software Foundation
The language behind most modern engineering computation, and its full ecosystem.
Get it from the vendor
wxMaxima — Symbolic Algebra
Maxima Project / wxMaxima
Graphical front end to the Maxima computer algebra system — derivatives, integrals and equations in mathematical notation.
Get it from the vendorBottom line
If what you are missing is Simulink, Scilab with Xcos is the first place to look. For general numerical computing, compare it with Octave and Python too.