[daip] Re: quick question- FITS IDI stokes parameters

[Eric Greisen]

Melissa Soriano writes:
 > 
 > Does the order of the STOKES parameters in the UV_DATA table's data matrix 
 > matter?  NO_STKS specifies the number of stokes parameter and STK_1 
 > specifies the first stokes parameter.  So we know which stokes parameter 
 > should go first in the data matrix STOKES axis.  But what about the order 
 > of the rest of the entries in the STOKES axis?  Can the parameters be 
 > entered in any order?  We are assuming that the order is arbitrary since no 
 > order is specified in the FITS-IDI documentation.
 > 

In this case, absolutely not.  STOKES is an arbitrary axis that has a
conventional meaning.  The standard values are

1  for I
2      Q
3      U
4      V

-1     RR
-2     LL
-3     RL
-4     LR

-5     XX
-6     YY
-7     XY
-8     YX

AIPS expects either to get I,Q,U,V in that order with a stokes axis
increment of +1 (it can handle some other cases sometimes) or

RR,LL,RL,LR in that order.

Occasionally an interferometer only has RR or LL and we get just that
one.

The IDI document does not specify what is now an IAU standard and was
well known at that time.

Eric Greisen