Hi Juergen.
I deployed telcal version 1.2.0 today to both my weekly and daily environments.

evla-telcal-1.2.0 (07dec2009):
Fixed Subscan row lookup in the SDM's SubscanTable.xml.
Fixed problem looking up band in Hz instead of MHz.
Changed the antsol amplitude calculation so it no longer divides by numChan.
Changed the antsol amplitude calculation so it no longer returns the sqrt of power (aka the voltageGain), but instead returns the amplitude as power (sqrt(r*r+i*i)).
Convert skyFreq from Hz to MHz before doing the pointing offset calculation.
Changed amplitude to amplitudeSquared in PointingData for Reference Pointing. We need to square the amplitude we get from antsol so that we end up with real*real + imag*imag.
Bdf file size changed from an int to a long in SdmInfo and subsequent use of this data.
Changed default subband for testing from 0 to 1 because subband 0 has boundary data ugliness.
Added ptgsol text file logger parameters to the telcal.properties and TelCalConstants.java files.
Updated the ptgsol.xml upper size limit (udp.maxPtgsolDatagramBytes) in telcal.properties.
Changed Reference Pointing to
1. Use evla-commons' ptgsol.xsd file again.
2. Collect and solve pointing solutions by IF rather than by polarization.
3. Only use the last 5 good positions to solve.
4. Added bad data checks and logging.
5. Solve if any of the four IFs have a valid set of data points.
Removed the setting of deprecated antsol.xsd subId field.
Made the following updates for Reference Pointing:
1. Created PeekTextFileWriter which will log all the details of the pointing calculations.
2. Changed the ptgsol multicast send method to only send if TelCalConstants.SEND_MULTICASTS is true.
Removed ability to print Antsol solutions to standard out when running TelCalStandAlone.
AntsolLogging is now integrated into TelCal and not a seperate process that reads the antsol multicasts.
TelCalStandAlone is now using the SdmMessages.tmp file to recreate the SdmInfo messages used to process the SDM. This is only a temp change and the old method remains in place for later use.
TelCal now uses the state field instead of subscanNumber for determining pointing position.
AntennaData.java is no longer dependent on the first subscan having valid data for all antenna. If valid antenna data doesn't come along until a subsequent subscan, that's fine.
I no longer pass hints between scans. It was causing problem finding solutions after a scan of bad data.


-- JuergenOtt - 2009-12-07

This topic: EVLA > WebHome > SoftwareTracking > EmailKC07DEC09
Topic revision: 2009-12-07, JuergenOtt
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding NRAO Public Wiki? Send feedback