AIPS: 1st Quarter 2003


Key Developments

1.           AIPS has been ported to the McIntosh OS/X operating system. At present, it appears that it is not making optimal use of the hardware, since the AIPSmark is lower than expected for the speed of the processor. Investigation of the tuning of the software for McIntosh has been halted, due to the lack of a modern McIntosh machine in the AOC.

2.           Various bug-fixes and improvements have been made in the task CONFI, which is used to optimize array configurations. Many of the changes are associated with improvements in the algorithm that searches for the worst sidelobe and modifies the configuration to attempt to reduce that sidelobe.

3.           UVSIM was modified to remove unused code, as well as changing the observation date and observing epoch to 1 Jan 2000. A true IAT time calculation was added based on the input coordinates of the observing array. The related task UVFIX was modified to make correct calculations for the GMRT, whose coordinate system (u,v,w) seems to be opposite from the VLA and other definitions. The investigation of the GMRT coordinate issue exposed various other bits of inefficient or erroneous code which were dealt with.

4.           A number of small modifications were made to the VLBA data-calibration utilities, both for clarity and to correct small bugs.

5.           Various changes were made to FILLM to make the reading of VLA data from disk a more robust process. Work began on a new real-time task, currently dubbed FIXXM, which is intended to be much more robust for real-time filling of VLA data. It is expected that this will be added to the system in the second or third quarter.

6.           Investigation into improved smoothing, blanking, and interpolation of calibration tables was begun. The code and the techniques are rather complicated, so completely new tasks have not yet been entered into the system.

7.           An investigation into possible improvements in phase calibration using externally supplied troposphere calibration was begun.

 

Goals for Second Quarter 2003

We anticipate a reduction in completion of new capabilities during the second quarter, due to a long vacation for the lead AIPS software developer. The following efforts are expected:

 

1.           Continue user support and bug fixes, as the major portion of AIPS effort.

2.           Add the more robust code for real-time filling of VLA data into the system.

3.           Continue investigation of improved smoothing and blanking in calibration.

4.           Continue the moderate-term project to explore improved troposphere calibration.

5.           Add Tsys values for a single VLA antenna into the calibration data supplied with VLBA correlator output.

6.           Test and enhance AIPS performance on McIntosh computers with the G4 processor (dependent on availability of a machine for testing.