[daip] IMAGR message. What does it mean?

[Eric Greisen]

Lincoln Greenhill writes:

 > I am running an 8192 image with about 300000 visibilities. When IMAGR
 > (31DEC99) reaches the "minimum clean flux density" message triggered by
 > FLUX, it fails to stop.  Instead, it starts incrementing the iteration
 > counter by one at a time.  It also prints the warning message, "CLMPAC:
 > NOT ENOUGH ROOM FOR ALL MAP POINTS".  The complete message text follows.  

         Why are you running 31DEC99?  And is it really called 31DEC99
or 31DEC00 which was the final version of that release.  31DEC01 is
available at your site and it has had issues of stopping on FLUX
addressed.

 > I note that I have been table to cut off IMAGR by communicating a small
 > value of NITER via the TELL facility.
 > 
 > IMAGR1: Reached minimum algorithm flux = 100.993 MilliJy iter=     2784
 > IMAGR1: Total Cleaned flux density    =  267.339      Jy     2784 comps
 > IMAGR1: ALGSUB: Ipol gridded model subtraction, chans    1 through    1
 > IMAGR1: Field   1 min = -512.8 MilliJy,max =  540.3 MilliJy
 > IMAGR1: BGC Clean: using  205 cell beam + residuals >    32.92 MilliJy
 > IMAGR1:    19992 Residual map points loaded
 > IMAGR1: Reached min. Clean flux density=  79.967 MilliJy iter=     3216
 > IMAGR1: Total Cleaned flux density    =  270.837      Jy     3216 comps
 > IMAGR1: ALGSUB: Ipol gridded model subtraction, chans    1 through    1
 > >IMAGR1: Field   1 min = -504.5 MilliJy,max =  534.5 MilliJy
 > IMAGR1: BGC Clean: using 1615 cell beam + residuals >   -61.04 MicroJy
 > IMAGR1: CLMPAC: NOT ENOUGH ROOM FOR ALL MAP POINTS
 > IMAGR1:      844 Residual map points loaded
 > IMAGR1: Reached min. Clean flux density=  74.887 MilliJy iter=     3217
 > IMAGR1: Total Cleaned flux density    =  270.844      Jy     3217 comps
 > IMAGR1: VISDFT: Begin DFT component subtraction
 > IMAGR1: I Polarization model processed
 > 
 > 
 > I have never seen this before.  Is the AP size too small?
 > RUN SETPAR shows:  
 > 
 > 29  Pseudo-AP 2nd memory (1024s)        12736   
 > 

I assume that is the actual size of the AP core and that is enough for
many more than 20000 pixels to clean (as the previous iteration
shows).  For some reason - perhaps because it should have quit -
some pointers are not right and it thinks the ap is fullwith only 844
map points - which would explain the clean finding only 75 mJy rather
than something closer to 80 (your cutoff I assume).  I have not seen
this error.  If you are actually running an old version, try the new
one.  If not, I am going to need to see your inputs et al. and maybe
Amy will need to try some imaging.

Eric