User Tools

Site Tools


Tips and Tricks

This page basically lists a couple of useful information to create good schedules

General tips

  • If you have troubles reading anything in the GUI simply change the window size.
  • Save your station network if you use it several times.
  • By clicking Load SCHED parameters you will always load the last used scheduling parameters.
  • :!: Always write log files. This makes it easier to understand current results and the Matlab Command Windows stays clean.
  • You do not need to display the sky coverage via the checkbox in the GUI. This plot is also saved in your output directory.
  • :!: Always make sure you use the newest catalog files.
  • :!: Use the mutli scheduling tool!
  • Simulate your results. A high number of observations does not generally mean the schedule is better!
  • optimization conditions can help avoid single baseline scans.
  • use the iterative approach for optimization conditons
  • :!: Always create and use a new source file unless you have a good reason not to do. This should be checked in the GUI all the times.
  • If something is not working or you have a question write us an email.
  • :!: If you use the multi scheduling tool, use parallel processing.

Good parameters

  • The weight factors in the param.txt file are the most important factors to create a good schedule
  • Trying different weight factors via the multi scheduling tool is very easy.
  • As a start use the following factors:
    • for a regional network like the Austral stations try
    • for a global network try:
    • only the relative ratio between the weight factors is important! (PARA.WEIGHT_NUMBER_OF_OBS 1, PARA.WEIGHT_SKY_COVERAGE 2, PARA.WEIGHT_SCAN_END_TIME 4 will give the same result as PARA.WEIGHT_NUMBER_OF_OBS 10, PARA.WEIGHT_SKY_COVERAGE 20, PARA.WEIGHT_SCAN_END_TIME 40)
  • If you have a global network use subnetting
  • If you have a regional network do not use subnetting
  • the minimum station number should be always 2, except you have good reasons
  • use the optimization conditions to avoid single baseline scans
  • Use PARA.FILLINMODE 1 unless you have a good reason to use 2 or 12. Using 0 is never a good idea.
  • Usually around 100 sources should be enough to create a schedule.
  • Check your Matlab Command Window (or body.txt) if you have any 0-scan subnets. If this happens you do not have enough sources. Try to run the schedule again with a lower minimum flux density or reduce the PARA.MIN_SRCRP value (it is maybe better to reduce the PARA.MIN_SRCRP value)
  • It might help to reduce the PARA.MAXSLEWTIME a bit, but it is generally better to simply increase the PARA.WEIGHT_SCAN_END_TIME instead
  • All other parameters are not very critical. Only make sure that you do not have to high conditions for this parameters. (like a PARA.MAX_WAIT of 10 would yield to a problem, or a PARA.MAXSLEWTIME 10… it is generally better to use high numbers here and not to constrain anything here)
  • Check if you have any 0-scan subnet in your Matlab Command Window or in your body.txt.
public/vievs_manual/scheduling/tips.txt · Last modified: 2017/06/10 09:00 by mschartn