[daip] AIPS problem

Sat May 25 05:07:55 EDT 2002

Dear Eric,

I may have encountered some possible bugs in the AIPS NEW and
AIPS TST version, which we use here at Bonn. 

1. There seems to be a problem with FITLD. Recent VLBA data which I
got on DDS3 (BK90B, Feb. 2002) could not be loaded with FITLD on our HP's
(using an HP tape drive, and a HP disk).
Together with W. Alef we tried various combinations of machines and
tapedrives. Finally we found that FITLD reads the data 
using AIPS on Linux PC and Linux tape drive. On the HP's FITLD always
crashed after reading the first file from tape. I skipped to another
file (eg. the 2nd, 3rd, etc.) but it also crashed there. The crash always
happend near the stop time of the data set, which we tried to read in.
Further below I show the detailed error message.

2. There is a plotting problem of POSSM on LINUX AIPS, if one tries 
to plot STOKES 'LR'. If I specify STOKES 'LL','RR','RL' it works.
If I specify polplot 'LR/RR' it also works. I used 
NPLOTS 9
APARM  1 0 0 0 0 0 1 0 1 0    

This happens for different data sets and also one other user (U. Bach)
had the same problem. 

For STOKES 'RL' I get on the message window:

pc134 > POSSM1: Task POSSM  (release of 31DEC02) begins
pc134 > POSSM1: Averaged       113 visibilities from total data set of     9628
pc134 > POSSM1: POSSUV: on baseline(s) BR       - LA
pc134 > POSSM1: Will plot    32 channels *  4 IFs polarization RL
pc134 > POSSM1: FILLPL: Amplitude mean:  2.0821E-04  rms:  1.0957E-04
pc134 > POSSM1: FILLPL: Phase mean:  3.5508E+01  rms:  1.0809E+02
pc134 > POSSM1: Doing: BR - LA     IF: 1
pc134 > POSSM1: -------------------------
pc134 > POSSM1: Averaged       114 visibilities from total data set of     9628
pc134 > POSSM1: POSSUV: on baseline(s) FD       - LA
pc134 > POSSM1: Will plot    32 channels *  4 IFs polarization RL
pc134 > POSSM1: FILLPL: Amplitude mean:  2.8771E-04  rms:  1.2934E-04
pc134 > POSSM1: FILLPL: Phase mean:  5.2018E+01  rms:  1.0778E+02
pc134 > POSSM1: Doing: FD - LA     IF: 1
pc134 > POSSM1: -------------------------

and the plots appear.

For STOKES 'LR' I get on the message window:
pc134 > POSSM1: Task POSSM  (release of 31DEC02) begins
pc134 > POSSM1: Averaged       113 visibilities from total data set of     9628
pc134 > POSSM1: POSSUV: on baseline(s) BR       - LA
pc134 > POSSM1: Averaged       114 visibilities from total data set of     9628
pc134 > POSSM1: POSSUV: on baseline(s) FD       - LA

no plot appears on the TV (or in a PL table).

best regards,
                   Thomas

This is the message and error from FITLD:

HP218 > FITLD3: Task FITLD  (release of 31DEC01) begins
HP218 > FITLD3: Reading tape drive number   1
HP218 > FITLD3: Reading tape at current position
HP218 > FITLD3: Correlator code version: 4.22
HP218 > FITLD3:     FITLD Version 10.5.1
HP218 > FITLD3:     Last updated on November 17 1999
HP218 > FITLD3:     Recognize and handle OVLBI time filters
HP218 > FITLD3:     Fix duplicate SU entries, result of CALCODEs
HP218 > FITLD3:     Read multiple uv tables per FITS file
HP218 > FITLD3:     Deal correctly with weights as an array
HP218 > FITLD3:     Antenna numbers in MC table numbered correctly
HP218 > FITLD3:     Read FLAG and WEATHER tables
HP218 > FITLD3:     Reads a-priori information for VLBA
HP218 > FITLD3:     Adjusts AN table reference frequency
HP218 > FITLD3:     Can now do calibration transfer
HP218 > FITLD3:     Checks for invalid calibration tables
HP218 > FITLD3:     Does not round frequencies excessively
HP218 > FITLD3:     Now concatenates cal transfer tables
HP218 > FITLD3:     Now handles correct u,v,w labels
HP218 > FITLD3:     Now handles explicit subarrays
HP218 > FITLD3:     Tolerates data out of time order within days
HP218 > FITLD3:     Weights now proportional to integration time for VLBA
HP218 > FITLD3:     VLBA saturation correction now applied correctly
HP218 > FITLD3: Data with weight <  0.65 will be permanently flagged
HP218 > FITLD3: Create BK90B       .FITLD .   1 (UV)  on disk  3  cno    9
HP218 > FITLD3: Sky frequencies (MHz) before re-ordering occurs:
HP218 > FITLD3: Incoming FREQID #   1
HP218 > FITLD3: 86210.49  86218.49  86226.49  86234.49
HP218 > FITLD3: Incoming FREQID #   2
HP218 > FITLD3: 43201.49  43209.49  43217.49  43225.49
HP218 > FITLD3: Incoming FREQID #   3
HP218 > FITLD3: 22217.49  22225.49  22233.49  22241.49
HP218 > FITLD3: Warning: table type OB is of zero length
HP218 > FITLD3: CL table interval set at  0.05 minutes
HP218 > FITLD3:            !!!!!!!!!!!
HP218 > FITLD3: Digital corrections being performed
HP218 > FITLD3: Total power corrected even if no zero-padding
HP218 > FITLD3:            !!!!!!!!!!!
HP218 > FITLD3: Current file has    47311 visibilities
HP218 > FITLD3: File ref. freq:    86210.49 MHz
HP218 > FITLD3: Found 1633+382         at   0/08:06:59 src #   1
HP218 > FITLD3:                                        fqid #   1
HP218 > FITLD3:    10000 vis. written
HP218 > FITLD3: UV table spanned time:   0/08:07:00 -  0/08:36:53

....

HP218 > FITLD3:    20000 vis. written
HP218 > FITLD3: Found 1633+382         at   0/09:25:48 src #   1
HP218 > FITLD3:                                        fqid #   3
HP218 > FITLD3:    30000 vis. written
HP218 > FITLD3: UV table spanned time:   0/09:06:54 -  0/09:28:40
HP218 > FITLD3:    31347 vis. written
HP218 > FITLD3: ZABORS: signal 11 received
HP218 > FITLD3: ABORT!

it fails:
on different HP machines 
on different tape drives
in versions NEW and TST
with and without freq. selection (SELFFREQ, FQTOL)
with and without compression (DOUVCOMP)   
on first or second or third file of DAT (DDS3)
if only one source is selected
it always crashes at the end of one file in standard
VLBA data distribution tape. Perhaps it has problems
writing the tables.

-- 
Thomas P. Krichbaum                              Tel.: (Germany +49)-228-525-295
Max-Planck-Institut fuer Radioastronomie         Fax : (Germany +49)-228-525-229
Auf dem Huegel 69                         e-mail:   tkrichbaum at mpifr-bonn.mpg.de
D-53121 Bonn / Germany                    www.mpifr-bonn.mpg.de/staff/tkrichbaum