Seminars Archive

Data Analysis and Optical Simulation within the Orange engine

Abstract
Orange is a software product by University of Ljubljana, providing a versatile environment for developers, researchers and data mining practitioners. Thanks to Python, a new generation scripting language and programming environment, your data mining scripts may be simple but powerful. To further allow for fast prototyping, Orange employs a component-based approach: you can implement your analysis method just like putting together the LEGO bricks, or even use an existing algorithm and replace some of its standard components with your own ones. What are Orange components to scripting are Orange widgets to visual programming. Widgets employ a specially designed communication mechanism for passing objects like data sets, attribute lists, learners, classifiers, regressors and alike, allowing to easily build rather complex data mining schemes that use state-of-the-art approaches and techniques. Orange is a comprehensive, component-based framework for machine learning and data mining. Orange is intended for both experienced users and researchers in machine learning who want to develop and test their own algorithms while reusing as much of the code as possible, and for those just entering the field who can either write short Python scripts for data analysis or enjoy a powerful while easy-to-use visual programming environment. Recently, the Orange engine has been successfully customized for scientific purposes, becoming an open Source graphical environment for simulating X-ray optics, called OASYS (OrAnge SYnchrotron Suite), thanks to the collaboration between Elettra-Sincrotrone Trieste, ESRF and University of Ljubljana. OASYS already contains, as a widget collection, optical simulation tools like ShadowOui, the evolution of the ray-tracing tool Shadow3, and XOPPY, the evolution of the optical calculation tool XOP. Instead, (Infra)Orange, an Orange-based data analysis tool for hyperspectral data, came to life from the collaboration between Elettra Sincrotrone Trieste, Soleil and University of Ljubljana, and it is going to be released. The seminar is divided in the following talks: "Orange: interactive data mining” by Tomaž Curk (University of Ljubljana); "(Infra)Orange: a new powerful tool for hyperspectral data analysis” by Giovanni Birarda (Elettra-Sincrotrone Trieste); "OASYS and ShadowOui: a graphic user interface for Shadow3 based on a development environment for synchrotron radiation simulation tools” by Luca Rebuffi (Elettra-Sincrotrone Trieste); "Simulation of new sources for the Extreme Brilliant Source at ESRF” by Manuel Sanchez del Rio (ESRF); "Data Analysis software developments at the ESRF” by Armando Solè (ESRF).