[daip] Re: CPASS AIPS bug

[Eric Greisen]

Wouter Vlemmings writes:

 > For some time now, it seems that CPASS has a serious bug in it.
 > No matter what input parameters are used, it calculated something and then 
 > end with:
 > 
 > cluste> CPASS1: BASOLV: SOLVER RETURNS ERROR    2
 > cluste> CPASS1: Appears to have ended successfully
 > 
 > producing an empty BP table. This is especially troubling for sensitive 
 > polarization observations. Any idea what is going wrong.
 > 

CPASS is a problematic task that seems to converge properly only when
the data are normalized extensively.  BPASSPRM(3) must be 0 and
BPASSPRM(5) too.   The number of terms may be significant.

BPASS is better for solving for an average gross shape of the bandpass
with CPASS then used to fit a nearly flat neraly constant function.

Eric Greisen