[daip] New Staff Reply - [!CEV-851818]: Help with AIPS UVRFI task

[Eric Greisen]

New Staff Reply: Help with AIPS UVRFI task

As I wrote before, UVRFI is a VERY experimental task and its
author has retired.  It now will try to handle TB data as well as BT
although the former will require vast amounts of (dynamic) memory.
That was one of the things I tried to put in during 2012 (but was unable
to test).  I am not certain that this task will get much attention unless
its real utility for some sort of data is demonstrated.  VLA data tends
to have RFI that does not have a fixed fringe frequency and so is less
amenable to the algorithm (moving sources such as satellites and airplanes).

But the issue seems to be with AVSPC in that the data are "normal"
going in to that task and not right after.  AVSPC averages those channels
that are good - so even if there was only 1 channel "good" in a group
of 16, you should get data with weight > 0 in that group.  Therefore
you should not get an increase in per cent of flagged channels, but a
decrease.  I have no real reason to suspect AVSPC of errors so I would
worry either that UVRFI has messed up some output parameters confusing AVSPC
or that the data going into AVSPC are really not right.

I will need you to look at these data more closely to see if we can debug the issue
or I suppose to put an illustrative piece of the data on an anonymous ftp
site where I can fetch it.

Eric Greisen


Ticket Details
===================
Ticket ID: CEV-851818
Department: AIPS Data Processing
Priority: Default
Status: Open
Link:  https://help.nrao.edu/staff/index.php?_m=tickets&_a=viewticket&ticketid=2610