Go to the Indico Home Page Oldest event Older event Up to category Newer event Newest event Printable version
  • Compact style
  • Indico style
  • Indico style - inline minutes
  • Indico style - numbered
  • Indico style - numbered + minutes
  • Indico Weeks View
    • Switch to the management area of this event
    Login

    Sixtrack postprocessing

    → Europe/Zurich
    Description

     

    https://cern.zoom.us/j/7288039887?pwd=aTNzbW9xWXRnMzNSam12YU9Ob1ZIUT09

    Hide

    Present (in zoom): Massimo, Frederik, Felix, Michael S., Kostas, Ewen, Joschua, Sofia, Loic, Giovanni I., Guido, Laurent, Riccardo.

     

    Massimo: Currently using sixtracklib with the only postprocessing of action calculation without taking coupling into account. The impact of coupling can be judged by looking at results and it is checked that results are not significantly affected.

    Recent developments no fully implemented in sixtrack/sixdesk, including full fix of the writebinl bug. Riccardo mentions that the fix of the writebinl bug in Sixtrack is not super easy. Ewen comments that impact of the bug can be minimized by reducing writebinl value.

    SixDB would need fix of the writebinl bug for the DA vs turns studies or at the source, in Sixtrack.

    BOINC needs the review of the assimilator plus improving the efficiency.

     

    Ewen: Action calculation is significantly distorted in presence of large coupling and non-linearities or in presence of a combination of various non-linear sources. MO or coupling alone do not seem to feature any issue.

    Various algorithms to compute actions are presented with pros and cons. Implementation is at the moment fully independent and it might be hard to have it robust. Might require being sure about tune finding. Simple implementations could still capture most of the correction to action.

    Kostas: Postprocessing sixtraklib implemented in python to compute T (or W) matrix and closed orbit via tracking. This calculation is the same as for Sixdesk/SixDB on the initial conditions. Riccardo asked for the code to be made available to sixtracklib for other users. Kostas and Gianni agree. Gianni also mentions that alternative easy implementation could be to take T matrix from MADX.

     

    Michael: Work on Pysixdesk by Loic and Michael.  Pysixdesk made compatible with DIST  + SIMU modules in Sixtrack. However Sixtrack version in BOINC does not include these modules. Alessio was planning to update sixtrack executable in ~June but left. Rogelio to ask Alessio about this update. Pysixdesk does not include any postprocessing at the moment and manpower would be needed for this.

     

     

     

    There are minutes attached to this event. Show them.