Visualization of Large Amount of Spectra in Virtual Observatory Environment

International Journal of Automation and Computing Visualization of Large Amount of Spectra in Virtual Observatory Environment Petr Saloun 1 David Andresic 1 Petr Skoda 0 1 Ivan Zelinka 1 0 GNU General Public License - www.gnu.org/licenses/gpl- 3.0.txt 1 Regular paper Special Issue on Recent Advances on Complex Systems Control, Modelling and Prediction II Manuscript received August 19, 2013; accepted October 11, 2013 This paper was supported by Agency of the Czech Republic-GACR (No. P103/13/08195S), the Development of Human Resources in Re- search and Development of Latest Soft Computing Methods and Their Application in Practical Project, Operational Programme Education for Competitiveness (No. CZ. 1.07/2.3.00/20.0072), ESF , State Budget of the Czech Republic , SGS (No. SP2013/114) , VSB- Technical University of Ostrava , Czech Republic This paper presents overview of new features so far prepared for new version of spectral analysis tool SPLAT-VO that allows to retrieve a large amount of spectra (and other data) based on its characteristics by detailed querying a virtual observatory s resources. The overview is focused on enhancements of user experience, work with simple application messaging protocol (SAMP) and other interoperability that improves work with global list of spectra, plot window and analysis menu. 3Standardized, coordinated and developed by international virtual observatory alliance (IVOA) - www.ivoa.net 4www.ivoa.net/Documents/SSA/ 5www.ivoa.net/Documents/latest/SAMP.html 6www.star.bristol.ac.uk/mbt/topcat/ 7www.g-vo.org 8www.asu.cas.cz 9VSB - Technical University of Ostrava, www.vsb.cz SPLAT-VO; virtual observatory; spectra; simple application message protocol (SAMP); flexible image transport system (FITS) Introduction The large amount of spectra requires capable tools for selecting by varied criteria, processing and visualization. Large amounts of data exist, and today s tools that the astronomers are using (like SPLAT-VO) need to be adjust accordingly. Virtual observatory (VO) is a collection of interoperating data archives and software tools which utilize the internet to form a scientific research environment in which astronomical research programs can be conducted. VO offers huge data-sets with thousands of fuzzy interconnected attributes for researcher, but it is necessary to select the proper one and show them. SPLAT is a shortcut for spectral analysis tool[1], and it is a tool for displaying, modifying and analysing astronomical spectra, for example. The examples can be seen in Figs. 1 3, respectively. SPLAT was developed in 2003 as a part of Starlink (and its STARJAVA package) project[1]. The Starlink was closed down in 2005 and took over by joint astronomy centre1 that relicensed some of its parts under GNU/GPL licence in 2006[2]2. During its development, the SPLAT was extended to include facilities that allowed an interoperability with the virtual observatory[1]3. The VO is a collection of astronomical archives and software tools that utilizes the internet to allow international collaboration and scientific research[3]. SPLAT itself uses the simple spectra access protocol (SSAP)4 for obtaining the spectra from SSAP servers and simple application messaging protocol (SAMP)5 to interoperate with other SAMP-compatible tools like tool for operations on catalogues and tables (TOPCAT)6. This paper completes our work, which is about to be accepted by developers and research community of SPLAT, our previous paper[4] described our first results in this area. Since 2012, SPLAT-VO has been developed by the Ger man Astrophysical Virtual Observatory (GAVO)7 in coop eration with the Astronomical Institute of the Academy of Sciences of the Czech Republic[1]8. Its development is cur rently focused on implementing new VO standards. The need to extend the scientific features that do not immediately relate to the VO and to fix some of the SPLAT s bugs or deficiencies resulted in a bachelor thesis at VSB-TUO9 supervised by Petr Skoda of Astronomical Institute of the Academy of Sciences of the Czech Republic. This thesis scoped some changes to SPLAT-VO as described below. In the near future, after dealing with coordination and unification of SPLAT s development, we intend to continue to add new features and implement new VO standards so that the SPLAT can be a feature-rich, user-friendly and even more portable scientific tool and reference implementation of new VO standards. Fig. 1 SPLAT-VO at work New features Besides the strictly virtual observatory related changes, which will allow the user even more detailed searching for spectra, the upcoming version of SPLAT-VO focuses on the user experience, interoperability with other tools and work with the spectra itself. On the next few pages, we describe these new features. We believe that all together will make your work with SPLAT more enjoyable. Plot window The purpose of plot window is quite simple: Plot every spectrum selected in the global list of spectra and allow the user to perform analysis on it. Every plot window can contain one or more spectra and when working with SPLAT, the user can open many plot windows with many spectra plotted in it. Added visual spectrum selection capability So far, when the user worked with multiple spectra in one common plot window, he has had a limited possibility to select just one spectrum for further actions. For example, selecting noisy spectrum and its deletion from plot window could be done only by trial-and-error procedure, and it is quite problematic and time-wasting. In the upcoming version, the user can click by mouse inside the plot window and SPLAT will automatically select the closest spectrum. The spectrum is selected in the local spectra drop-down menu above the plot itself and the global list of spectra in the main window. From there, it can be easily deleted or modified for plotting and analysis. The actual selection is done by getting the [x, y] coordinates of mouse click, which are then transformed to coordinates with the plot s [0, 0] as beginning. Then, the algorithm searches for all spectra with x coordinates containing the x coordinate of the mouse click. For this, the native (platform dependent) AST library10 from the original Starlink project is used. Each spectrum is represented by series of [x, y] coordinates, each coordinate represents one value. The AST library transforms this discrete list of values to a continuous line that can be seen in the plot window. This set of transformed coordinates is used by the selection algorithm. So when any spectrum is detected on the x coordinate, the algorithm will again use the modified binary search algorithm11 to select the closest spectrum on the y axis. This is quite similar to the procedure on x axis, the coordinates (transformed via AST) are used to locate the closest match of each x-colliding spectrum s [x, y] coordinates to the mouse click coordinates. The selection of the spectrum (...truncated)