[daip] spectral indices maps
Eric Greisen
egreisen at nrao.edu
Tue Jan 28 16:24:31 EST 2014
Maciej.Ceglowski at astri.uni.torun.pl wrote:
> I have problem with task 'rstor'.
>
> I want to create spectral indices map.
> I use 'comb' to produce 'NULL' map. Then I want to restore cc components
> from map that I have just "sum" in task 'comb'. When I run task I got:
>
> RSTOR1: CMPCRM: NO POINT CLEAN COMPONENTS FOUND
>
> I don't understand this. There are cc - components. Whole message:
>
> MACIEK> RSTOR1: Task RSTOR (release of 31DEC13) begins
> MACIEK> RSTOR1: Create J1254+08 .IRST . 2 (MA) on disk 1 cno 102
> MACIEK> RSTOR1: Copied CC file from vol/cno/vers 1 97 1 to 1 102 1
> MACIEK> RSTOR1: RSTOR / INPUT MAP ='J1254+08 ' . 'NULL ' . 2 1
> MACIEK> RSTOR1: RSTOR / CC MAP = 'J1254+08 ' . 'RRCLN ' . 2 1
> MACIEK> RSTOR1: RSTOR / RESTORED MAP ='J1254+08 ' . 'IRST ' . 2 1
> MACIEK> RSTOR1: RSTOR BLC = 1 1 1 1 1 1 1 / Plane
> CLEANed
> MACIEK> RSTOR1: RSTOR OUTVER = 1 /CLEAN comp. file version no.
> MACIEK> RSTOR1: RSTOR NITER = 9 /No. components restored.
> MACIEK> RSTOR1: RSTOR BMAJ= 0.00228 BMIN= 0.00101 BPA= 0.522 /CLEAN
> beam
> MACIEK> RSTOR1: QINIT: did a GET of 5120 Kwords, OFF 17520790895467
> MACIEK> RSTOR1: CMPCRM: NO POINT CLEAN COMPONENTS FOUND
> MACIEK> RSTOR1: Destroyed 1 extension files of type HI
> MACIEK> RSTOR1: Destroyed 1 extension files of type CC
> MACIEK> RSTOR1: Destroyed MA image file: catno= 102 disk= 1
> MACIEK> RSTOR1: Purports to die of UNNATURAL cause
Use PRTAB on the CC file you are using. RSTOR looks at the type column
and if it is not 0 it ignores the component. The error message
indicates that it found no components of type 0 - either because they
are extended or because there are no components.
The standard way to make spectral index images these days is different.
Make a spectral cube with IMAGR, then run CONVL on it to convolve all
planes to the resolution of the lowest frequency plane. Finally run
SPIXR on the CONVL output. This keeps the residual of each image at the
desired resolution, not just the flux in the components. You can also
build a cube with FQUBE and CONVL and SPIXR it if your images are at
rather different frequencies. Note that this requires a relatively
modern IMAGR that makes CG tables (to show what the resolution of each
plane in the cube actually is).
Eric Greisen
More information about the Daip
mailing list