Open source software for numerical computation

Skip to main content

Final review of Scilab projects in the Google Summer of Code 2012

Involved for the fourth consecutive year in the Google Summer of Code, Scilab and its community have mentored 10 students in this 2012 edition. 

All development work that has been performed on Scilab exceeded our expectations and we renew our thanks and congratulations to the students.

gsoc2012_summit

At the end of the Google Summer of Code, Scilab Enterprises has participated on 20 and 21 of October to the Mentor Summit at Google headquarters in Mountain View, California. Bringing together the mentors of each project, it is about 300 software experts from the Open Source World who exchanged on the latest developments of their fields and their experiences. The 2011 Summit allowed, for example, federating the development efforts between Octave, Scilab and Debian on the arpack-ng library (used to calculate the eigenvalues of large problems) based on arpack which was no longer maintained at that time.

Improved graphic controls by Marcos Cardinot

Marcos Cardinot has greatly improved graphics functionalities in Scilab. Thus, the future Scilab 5.5.0 version will include a new graphical editor (ged). Natively integrated in the graphics engine and in Scilab interface, these developments participate in improving the software functionalities, the graphics performances stability and facilitate the component maintenance.

gsoc2012_1

Easy and Interactive Plot by Caio Lucas dos Santos Souza

Caio and Pedro Lucas dos Santos Souza improved general interactivity of graphics by adding, for example, copy/paste in figures and editing legends or axis titles on the fly.

gsoc2012_2

Various improvements for the Scilab graphic features by Gustavo Libotte

Gustavo Libotte worked on improving the integrated datatips in Scilab. Now available natively, they are available through a dedicated field in the graphics handles. The interactivity is also greatly improved.

gsoc2012_3

Those three projects were dealing with Scilab graphics improvements. They were realized by students of the Polytechnic Institute of Brazil, mentored by Ricardo Fabri and Sylvestre Ledru. All the developments around the graphic will be included in Scilab 5.5.0 version.

Accurate Probability Distribution Function by Prateek Papriwal

Prateek Papriwal contributed to distfun module, which provides probability distribution functions used in probability problems. He contributed to the creation of numerically accurate functions for seven probability distributions (e.g. the Binomial distribution), which involves more than 20 different functions. His work improved the accuracy and robustness of the functions, their testing, and their help pages.

Integration of a "Design of experiments" module in Scilab by Maria Christopoulou

Scidoe module for planning experiments enables the generation of a set of points for the learning of certain types of models:

  • Linear models,
  • Polynomial models,
  • Non-linear models.

The types of plans included in the module are:

  • Latin Hypercube,
  • Central Composite,
  • Factoriel,
  • D-Optimal.

The module “Design of experiments” was developed by Maria Christopoulou with the participation of Michaël Baudin, Yann Collette and Per A. Brodtkorb.

Signal Processing Blockset by Sumeet

Sumeet developed a module which adds new blocks for signal processing to Xcos.
The following blocks are available:

  • FFT: frequency representation of a signal,
  • FIR: Finite Impulse Response Filter,
  • IIR: Infinite impulse response Filter,
  • BufferVect: transforming a time signal in a vector to allow matrix calculations,
  • max, min, rms, moving average, standard deviation.

This module was developed by Sumeet with the participation of Yann Collette.

SIP (Scilab Image Processing Toolbox) by Pallan Madhavan

Pallan added Leptonica support to SIP. Open source image processing and image analysis library written by Google, Leptonica focusses on automatic processing, cleaning, cropping, segmenting, and recognizing document images scanned with a camera or a scanner. Pallan worked on understanding and experimenting with the best document processing functionality from Leptonica and make it available in SIP.
From Scilab, users can now easily program and prototype document understanding solutions. This work opens many possibilities for producing better and automatic solutions for scanning and indexing documents and offline content for better search and discover ability.

Aerospace Blockset for Xcos by Paweł Zagórski

Paweł Zagórski has implemented aerospace dedicated blockset to Xcos. His work has first begun by implementing a representative Xcos test case to simulate and visualize satellite position on the Earth orbit. After this successful demo, he had also to design two more test cases: calculate access times between ground station and the satellite and the simulation of a weather balloon capsule falling from a high altitude.

gsoc2012_4

All this work has needed strong interaction between different Scilab and external communication channels (forge, atoms, Youtube, etc.). Paweł has also been mentored by two development teams: Scilab and Celestlab.

Introduce a REST client in Scilab by Rohan Kulkarni

Rohan work aimed at introducing a REST Client available from Scilab engine. Scilab can now make query on remote data sources and interact with other services via the REST protocol.

Scilab for Android by Guillaume Mazoyer

Guillaume Mazoyer worked on porting Scilab to Android. Although the project has not resulted in a final version of Scilab under Android, the work of Guillaume has set a number of critical issues, and lays the required initial groundwork for this future version.

We are looking for volunteers to complete the work.