; Load the file.
filein,"data/TGBT17A_506_11.raw.vegas"
; Display a summary of its contents.
summary
; Scans 14 through 26 contain mapping observations,
; with scan 27 being observations of a reference position.
; We will calibrate the mapping scans using scan 27 as
; reference.
; We want to save our data to a new SDFITS file so
; we can create a cube using the gbtgridder after calibration.
; If the file exists GBTIDL will append to it by default.
; Force the removal and creation of a new file with the new flag.
fileout,"outputs/NGC6946_cal_pol_0.fits",/new
; Loop over scans calibrating each one of them.
; Use the keepints option so that each integration is
; saved to the output.
for s=14,26,1 do begin & getsigref,s,27,ifnum=0,plnum=0,/keepints & endfor
; Now the orthogonal polarization.
; Keep them separated to inspect them independently.
fileout,"outputs/NGC6946_cal_pol_1.fits",/new
for s=14,26,1 do begin & getsigref,s,27,ifnum=0,plnum=1,/keepints & endfor
; After this it is possible to grid the data into a cube using:
; gbtgridder -v 5 outputs/NGC6946_cal_pol_{0,1}.fits
; This will create a cube named: NGC6946_1420_MHz_cube.fits
; Notice that the results are worse than using the gbtpipeline
; partially because the gbtpipeline will get rid of blanked
; integrations (all NaN rows), whereas GBTIDL keeps them in the output.