[daip] calib question

[Eric Greisen]

bob at physics.ucdavis.edu wrote:
> hi, so i have combined the c and b config data with reway run.
> when i try to selfcal, the P only works fine but when i try to
> selfcal both p and A it always fails.
> 
> CALIB6: Field  123 used       0 CCs
> CALIB6: Determining solutions using full vector averaging
> CALIB6: Writing SN table    2
> CALIB6: RPOL, IF= 1 The average gain over these antennas is  1.000E+00
> CALIB6: RPOL, IF= 2 The average gain over these antennas is  1.000E+00
> CALIB6: RPOL, IF= 3 The average gain over these antennas is  1.000E+00
> CALIB6: RPOL, IF= 4 The average gain over these antennas is  1.000E+00
> CALIB6: LPOL, IF= 1 The average gain over these antennas is  1.000E+00
> CALIB6: LPOL, IF= 2 The average gain over these antennas is  1.000E+00
> CALIB6: LPOL, IF= 3 The average gain over these antennas is  1.000E+00
> CALIB6: LPOL, IF= 4 The average gain over these antennas is  1.000E+00
> CALIB6: ERROR: NO VALID SOLUTIONS FOUND
> CALIB6: Destroyed UV image file: catno=      6 disk= 2
> CALIB6: Purports to die of UNNATURAL causes
> CALIB6: first 31DEC11 TST: Cpu=   2137.1  Real=   2451  IO=     27826
> 
> the inputs for calib are as follows. do you see any obvious mistakes?

> AIPS 6: WTUV          0                    Weight outside UVRANGE 0=0.
> AIPS 6: WEIGHTIT      0                    Modify data weights function
> AIPS 6:                                    Cal. info for input:

two things may help

set wtuv=0.01

try weightit=1 or even 2

The latter is because with heavily resolved sources the weight over all 
of the outer baselines becomes very small wrt the weight at short 
spacings in the divided data set used to solve for gains.  That 
discordance often leads to solution failure.

Eric