[daip] aips_gripe
Leonia Kogan
lkogan at aoc.nrao.edu
Wed Nov 14 15:26:01 EST 2001
Lincoln,
XMOM
OK.
It looks like I understand the problem now. I'll start thinking about
solving it inside of XMOM. I am going to start with constant size of
the box for all pixels and all velocities.
COMB
I am going to add an option to 'SUM':
IF ((MAP(1) .EQ. BLANK) .AND. (MAP(1) .EQ. BLANK)) THEN
OUTPUT = BLANK
ELSE
IF (MAP(1) .EQ. BLANK) MAP(1) = 0
IF (MAP(2) .EQ. BLANK) MAP(2) = 0
OUTPUT = A(1)*MAP(1) + A(2)*MAP(2) + A(3)
END IF
Leonia
----- Begin Included Message -----
>From lincoln at play.harvard.edu Wed Nov 14 11:52 MST 2001
Date: Wed, 14 Nov 2001 13:52:11 -0500 (EST)
From: Lincoln Greenhill <lincoln at play.harvard.edu>
To: Leonia Kogan <lkogan at zia.aoc.NRAO.EDU>
cc: daip at cv3.cv.nrao.edu
Subject: Re: [daip] aips_gripe
In-Reply-To: <200111141845.LAA15154 at bonito.aoc.nrao.edu>
MIME-Version: 1.0
X-Lines: 36
Status: RO
Hi Leonia,
> Now I start understanding that I has not understood the problem.
>
> So you want XMOM to calculate different threshold for different box
> at each velocity plane.
>
Yes, I think that ONE way to solve the problem I have described
(i.e., position-dependent dynamic range in each plane of a cube)
is for XMOM to use a position dependent threshold. One way to
do this is to draw a box around each pixel, compute and RMS, and
determine whether the pixel has a flux that is above or below
a threshold of N-sigma.
> The boxes positions and sizes for each plane should be given by the user.
The size of the box is given by the user. Remember, this is a moving
box, not a static one (i.e., not like the box in IMSTAT).
> If YES, YES then
> I have understood the problem.
> It looks like this problem solving is a subject of
> other task before using XMOM. This task should blank the qube
I think that XMOM can logically perform the operation I describe.
However, it does not matter to me. The operation could be performed
in BLANK (e.g., an OPCODE like IN2C).
Regards,
Lincoln
----- End Included Message -----
More information about the Daip
mailing list