MOSSDI fails on large mosaic?

Is MIRIAD being a pain? Let us know your experience.

Moderator: Mark.Wieringa

MOSSDI fails on large mosaic?

Postby and460 » Wed Feb 05, 2014 2:20 pm

Hi guys,

I am trying to MOSSDI a large mosaic (340 pointings!). It seems to start going just fine, but whenever it gets to a certain residual maximum flux density level (9.090E-01 Jy), it chokes, and just endlessly iterates with no change in the max or min flux levels. Are you able to give me any idea what may be going on here?

To demonstrate the problem, here is the output of two runs, the first with a loop gain of 0.25 and the second with 0.1:

mossdi% inp
Task: mossdi
map = /test.imap
beam = /test.ibeam
model =
out = /test.mossdi_i
gain = 0.25
niters = 10000
cutoff = 0.010
clip =
region =
options =
mossdi% go

mossdi: Revision 1.8, 2011/10/31 00:37:30 UTC

Beginning to CLEAN ...
Steer Iterations: 9
Residual min,max,rms: -2.728E-01 1.467E+00 1.437E-02
Total CLEANed flux: 5.202E-01
Steer Iterations: 17
Residual min,max,rms: -2.266E-01 1.112E+00 1.359E-02
Total CLEANed flux: 9.096E-01
Steer Iterations: 24
Residual min,max,rms: -2.266E-01 1.112E+00 1.313E-02
Total CLEANed flux: 1.201E+00
Steer Iterations: 33
Residual min,max,rms: -1.888E-01 9.090E-01 1.282E-02
Total CLEANed flux: 1.495E+00
Steer Iterations: 39
Residual min,max,rms: -1.888E-01 9.090E-01 1.255E-02
Total CLEANed flux: 1.713E+00
Steer Iterations: 48
Residual min,max,rms: -1.888E-01 9.090E-01 1.237E-02
Total CLEANed flux: 1.934E+00
Steer Iterations: 60
Residual min,max,rms: -1.888E-01 9.090E-01 1.210E-02
Total CLEANed flux: 2.259E+00
Steer Iterations: 73
Residual min,max,rms: -1.888E-01 9.090E-01 1.194E-02
Total CLEANed flux: 2.515E+00
^CMiriad shell cannot be interrupted, type 'help miriad' for info
mossdi%


mossdi% inp
Task: mossdi
map = /test.imap
beam = /test.ibeam
model =
out = /test.mossdi_i
gain = 0.1
niters = 10000
cutoff = 0.010
clip =
region =
options =
mossdi% go

mossdi% go

mossdi: Revision 1.8, 2011/10/31 00:37:30 UTC

### Warning: Error opening /test.mossdi_i, in XYOPEN
### Fatal Error: File exists
mossdi% go

mossdi: Revision 1.8, 2011/10/31 00:37:30 UTC

Beginning to CLEAN ...
Steer Iterations: 9
Residual min,max,rms: -3.285E-01 1.757E+00 1.512E-02
Total CLEANed flux: 2.081E-01
Steer Iterations: 18
Residual min,max,rms: -2.951E-01 1.583E+00 1.466E-02
Total CLEANed flux: 3.954E-01
Steer Iterations: 26
Residual min,max,rms: -2.648E-01 1.425E+00 1.428E-02
Total CLEANed flux: 5.636E-01
Steer Iterations: 34
Residual min,max,rms: -2.376E-01 1.282E+00 1.396E-02
Total CLEANed flux: 7.150E-01
Steer Iterations: 42
Residual min,max,rms: -2.266E-01 1.154E+00 1.370E-02
Total CLEANed flux: 8.512E-01
Steer Iterations: 49
Residual min,max,rms: -2.266E-01 1.112E+00 1.348E-02
Total CLEANed flux: 9.733E-01
Steer Iterations: 59
Residual min,max,rms: -2.177E-01 1.053E+00 1.320E-02
Total CLEANed flux: 1.160E+00
Steer Iterations: 75
Residual min,max,rms: -1.997E-01 9.363E-01 1.295E-02
Total CLEANed flux: 1.370E+00
Steer Iterations: 89
Residual min,max,rms: -1.888E-01 9.090E-01 1.273E-02
Total CLEANed flux: 1.556E+00
Steer Iterations: 100
Residual min,max,rms: -1.888E-01 9.090E-01 1.255E-02
Total CLEANed flux: 1.720E+00
Steer Iterations: 117
Residual min,max,rms: -1.888E-01 9.090E-01 1.239E-02
Total CLEANed flux: 1.903E+00
Steer Iterations: 130
Residual min,max,rms: -1.888E-01 9.090E-01 1.223E-02
Total CLEANed flux: 2.080E+00
Steer Iterations: 154
Residual min,max,rms: -1.888E-01 9.090E-01 1.209E-02
Total CLEANed flux: 2.284E+00
and460
 
Posts: 79
Joined: Thu Sep 01, 2011 12:23 pm

Re: MOSSDI fails on large mosaic?

Postby Mark.Wieringa » Wed Feb 05, 2014 3:24 pm

Hi Craig,

not exactly sure what is going on, but I notice that the total flux is still increasing and the rms decreasing, which looks promising. Maybe these max residuals are somewhere on a beam edge that doesn't ever get picked for cleaning. MOSSDI doesn't just pick the max residual like clean, it weighs the pixels by the sensitivity in the mosaic.

You could set the iteration count lower and have a look at the images that come out to see if that helps to figure out what is happening. You can restart the clean with the model you got out, if you don't want to start from scratch each time.

Cheers,

Mark
Mark.Wieringa
ATCA Expert
 
Posts: 284
Joined: Mon Feb 08, 2010 1:37 pm


Return to MIRIAD

Who is online

Users browsing this forum: No registered users and 1 guest

cron