Scilab

Scilab
Name	Scilab
Latest release	6.1.1
Developer	INRIA; Scilab Enterprises; Scilab Consortium
Written in	C; Fortran; C++
Operating system	Windows; Linux; macOS
License	CeCILL; GPL
Website	scilab.org

Scilab is an open-source numerical computation environment originally developed for numerical analysis, signal processing, control design, and data visualization. It was created by researchers at INRIA and later maintained by a consortium including Scilab Enterprises, and is used alongside environments such as MATLAB, GNU Octave, Python (programming language), and R (programming language). The system integrates with scientific projects and institutions including CNRS, University of Cambridge, Massachusetts Institute of Technology, École Polytechnique, and Technical University of Munich.

History

Scilab's origins date to research groups at INRIA in the early 1990s, influenced by numerical software like EISPACK, LINPACK, BLAS, and scholarly work from Jack Dongarra and collaborators. Its development overlapped with initiatives at University of Paris-Sud and collaborations with industry partners such as Schneider Electric and Thales Group. The project governance transitioned through entities including Scilab Consortium and commercial stewardship by Scilab Enterprises, with version releases coordinated alongside events at ACM SIGPLAN and IEEE Control Systems Society conferences. Funding and contributions came from European initiatives such as Horizon 2020 and national research agencies including ANR (France), while academic adoption expanded at institutions like Imperial College London, Delft University of Technology, EPFL, and Politecnico di Milano.

Features and Architecture

Scilab provides a matrix-oriented language with built-in libraries for linear algebra leveraging LAPACK and ARPACK, signal processing influenced by algorithms from FFTW and Discrete Fourier Transform methods. The graphical subsystem supports plotting comparable to toolsets in GNUplot and visualization frameworks used by ParaView and VTK. Integration layers include interfaces to Java (programming language), C (programming language), Fortran, and bridging projects connecting to Python (programming language) via gateways akin to NumPy interoperability. Scilab's plugin system and dynamic loading mimic architectures found in Eclipse (software) and GNOME. The build system and packaging draw on technologies from CMake, Autotools, and continuous integration practices seen on Jenkins and Travis CI.

Programming Language and Syntax

The Scilab language offers vectorized operations similar to those in A. R. Forsyth-style numerical languages and recreational derivatives related to Fortran 77 and MATLAB. Its scripting and function definitions resemble constructs taught at Massachusetts Institute of Technology and Stanford University numerical courses and follow conventions used in textbooks by Gene H. Golub, Charles F. Van Loan, and Gilbert Strang. Error handling and debugging utilities borrow paradigms from GNU Debugger workflows and unit testing practices comparable to JUnit or pytest when interfaced via wrappers. Interoperability with compiled code is enabled through mechanisms like the same foreign function interfaces used by R (programming language) packages and Julia (programming language)'s ccall patterns.

Toolboxes and Modules

Scilab's ecosystem includes modules for control systems inspired by algorithms standardized by IEEE, model-based design tools analogous to Simulink and interfaces with modeling languages such as Modelica. Signal processing toolboxes implement filters and transforms described in works by Alan V. Oppenheim and Ronald W. Schafer. Numerical optimization modules incorporate routines related to methods from Nelder–Mead and Levenberg–Marquardt algorithms cited in literature by Donald Knuth and John N. Hooker. Additional modules exist for image processing influenced by OpenCV conventions, parallel computing interfaces reflecting paradigms from MPI and OpenMP, and hardware interaction layers compatible with Arduino and Raspberry Pi ecosystems. Community-contributed toolboxes address domains covered by organizations like European Space Agency, CERN, and NASA.

Development and Community

Development has been driven by contributors from research centers including INRIA, CNRS, CNR, and companies such as EDF (Électricité de France), Airbus, and Bosch. Governance and release coordination have involved stakeholders from European Commission research programs and standards bodies like ISO. Community activities take place on platforms resembling those used by GitHub, mailing lists modeled after GNU projects, and conferences following formats of FOSDEM and SciPy community events. Educational adoption includes coursework at University of Oxford, University of California, Berkeley, National University of Singapore, and partnerships with organizations such as IEEE and ACM for workshops.

Reception and Use Cases

Scilab has been adopted in academia and industry for tasks similar to those performed with MATLAB at institutions like Princeton University and University of Tokyo, and in industrial settings at Siemens, Thales Group, and Schneider Electric for control design, signal processing, and prototyping. Comparative studies in journals such as those published by Elsevier and Springer have evaluated Scilab against MATLAB, GNU Octave, Python (programming language), and R (programming language). Use cases include numerical simulations in projects affiliated with European Space Agency, data analysis in research at Max Planck Society, and educational laboratories at ETH Zurich and Tsinghua University.

Category:Numerical software