[daip] imagr failure

[Eric Greisen]

bob at physics.ucdavis.edu wrote:
> good news. i ran ubavg and decreased the size by 3 and now it is running.
> maybe it was a fluke. i don't know.
> 
> 
> bob
> 
>> bob at physics.ucdavis.edu wrote:
>>>  hi, this was running on a really big uv file
>>>
>>>
>>>
>>> and it died as follows right after the very first round of cleaning
>>>
>>>
>>> IMAGR3: BGC Clean: using 1003 cell beam + residuals >   469.85 MicroJy
>>> IMAGR3:     4036 Residual map points loaded
>>> IMAGR3: Field   12 min allowed flux  =    5.193 MilliJy iter=       19
>>> IMAGR3: Field   12 Clean flux density=   29.923 MilliJy       19 comps
>>> IMAGR3: Total Cleaned flux density     =   29.923 MilliJy       19 comps
>>> IMAGR3: SETDO3: imaging done with multiple tangent planes
>>> IMAGR3: ALGMEM field   12 Ipol gridded model subtraction chns    1-  256
>>> IMAGR3: ALGMEM:  1024 - 0 cells, with     1449260 Pts
>>> IMAGR3: OUVIMG: MAKING IMAGE CLEAN field number 0001
>>> IMAGR3: OUVIMG: FROM UVdata work object
>>> IMAGR3: CLNUV2 : ERROR CLEANING CLEAN process object
>>> IMAGR3: CLNUV : ERROR CLEANING CLEAN process object

I think I have found an error related to the ONEBEAM true option.  The 
problem is that it should not have struck when doing field 1, just any 
other.  It is good that you have a work around.

Note that IMAGR offers the option of doing the UBAVG on the fly as the 
data are read to make the work file.  IM2PARM(11), (12), (and (13) for 
channels) implement this and it is quite efficient.

Eric Greisen