;+ ; NAME: ; RELVIS ; PURPOSE: ; Determines relative visibility, V(S)/V(0), where V(s) is the visibility ; as a function of fringe spacing s, and V(0) is the visibility at zero ; fringe spacing, i.e. total power. The relative visibility is independent ; of calibration, and will be unity for a point (unresolved) source but ; will decrease with increasing s for a resolved one. ; CATEGORY: ; OVRO APC ANALYSIS ; CALLING SEQUENCE: ; rv = relvis(filename[,tp][,tpmin=tpmin]) ; INPUTS: ; filename the name of the file containing output data from ANALYZE ; *after* preflare subtraction. If preflare subtraction ; has not been done, the results will be invalid. ; OPTIONAL (KEYWORD) INPUT PARAMETERS: ; tpmin if set, gives the minimum total power value for which to ; determine the relative visibility. If omitted, 5 sfu is ; the minimum value. ; ROUTINES CALLED: ; OUTPUTS: ; rv the relative visibility of size (NBL,NF,NTIMES) for each ; baseline. It will contain NaN where the total power is ; less than TPMIN. ; tp optional output containing the calibrated total power ; used in determining RV, of size (NANT,NF,NTIMES). ; COMMENTS: ; Because of the strange way that OVSA measures polarization, the ; strict relative visibility for circular polarization involves a ; fudge factor ranging from 0 to 1, depending on the polarization ; of the source. Write polarization measurements on baselines as Rij ; and Lij, and linear polarization measurements as Iij, where ij are ; antenna numbers. Write measurements in total power as Ai, where Ai ; are in general (Ri + Li), but for some antennas R and/or L are not ; measured at all. Thus, what is actually measured in total power is: ; Antenna RCP position LCP position ; A1 I1 = (R1 + L1) L1 ; A2 R2 I2 = (R2 + L2) ; An>2 In = (Rn + Ln) In = (Rn + Ln) ; The corresponding baseline correlated amplitudes are ; Baseline RCP position LCP position ; A12 sqrt(R1*R2) sqrt(L1*L2) ; A1n sqrt(I1*In) sqrt(L1*Ln) ; A2n sqrt(R2*Rn) sqrt(I2*In) ; others sqrt(Im*In) sqrt(Im*In) ; The fudge factors for the relative visibilities will then be ; (from, e.g. RV = Aij/sqrt(Ai*Aj)) ; Baseline RCP position LCP position ; A12 1/sqrt(1+L1/R1) 1/sqrt(R2/L2+1) ; A1n 1 1/sqrt(Rn/Ln+1) ; A2n 1/sqrt(1+Ln/Rn) 1 ; others 1 1 ; These fudge factors range from 0 to 1, depending on the polarization ; of the source, as follows: ; __Unpol'd__ _100% RCP_ _100% LCP_ ; BL RCP LCP RCP LCP RCP LCP ; A12 0.707 0.707 1 0 0 1 ; A1n 1 0.707 1 0 1 1 ; A2n 0.707 1 1 1 0 1 ; oth 1 1 1 1 1 1 ; ; The relative visibility of a point source would then have these values, ; and in order to convert this to unity, as required for a point source, ; the RV will be divided by these values. Note that RV is only useful for ; simple sources, so we will make the approximation that the total power ; polarization is the true source polarization. ; SIDE EFFECTS: ; RESTRICTIONS: ; MODIFICATION HISTORY: ; Written 04-Aug-2001 by Dale E. Gary ; 21-Nov-2001 DG ; Fixed a couple of errors. ; 27-Dec-2001 DG ; Removed subtraction of preflare, since it can be done much better ; in OVSA_PRESUB. ;- function relvis,file,tp,tpmin=tpmin if (n_elements(file) eq 0) then file = dialog_pickfile(path=!defaults.workdir,filter='*.sav') if (file eq '') then return,-1 restore,file ntimes = n_elements(tavg) nf = n_elements(f) nant = refcal.nant ; Number of antennas nbl = refcal.nbl ; Number of baselines if (size(avg,/n_dim) eq 3) then npol = 1 else npol = refcal.npol ; Number of polarizations in data reffac = (*refcal.pfactors) ; Extract REFCAL amplitude/TP factors antstr = ['1','2','4','5','6','7'] ; Apply calibration, and keep only TP part cal = apply_cal_all(tavg,avg,f,refcal) if (npol gt 1) then begin tpc = cal[0:nant-1,*,*,2] ; Save calibrated I TP tp = avg[0:nant-1,*,*,2] ; Uncalibrated tprc = reform(cal[1,*,*,0]) ; For R poln, save R for ant 2 in TP tplc = reform(cal[0,*,*,1]) ; For L poln, save L for ant 1 in TP endif else begin tpc = cal[0:nant-1,*,*] ; If only 1 poln, assume I tp = avg[0:nant-1,*,*] ; Uncalibrated endelse ; Now determine TP polarization if NPOL > 1 if (npol gt 1) then begin endif y = indgen(nbl)*2 + nant ; Channel numbers of sin channels x = y + 1 ; Channel numbers of cos channels amp = sqrt(avg[x,*,*,0]^2 + avg[y,*,*,0]^2) ; Amplitudes of baselines ; Generate a parallel array to the Amplitude array, with the quantity ; sqrt(T_i*T_j) for baseline i-j. tpbl = fltarr(nbl,nf,ntimes) tpblsfu = tpbl ; Parallel array to TPBL, but calibrated in SFU k = 0 bl_str = strarr(nbl) ants = [1,2,4,5,6,7] for iant = 0, nant-2 do for jant = iant+1, nant-1 do begin tpbl[k,*,*] = sqrt(tp[iant,*,*]*tp[jant,*,*]) tpblsfu[k,*,*] = sqrt(tpc[iant,*,*]*tpc[jant,*,*]) bl_str[k] = string(ants[iant],ants[jant],format='(I1,"-",I1)') k = k + 1 endfor ; Simple calculation to get Relative Visibility! rv = amp/tpbl ; When TPBL is too small, RV will not be significant, so blank values ; where the total power is less than about 5 SFU bad = where(tpblsfu lt 5.0,nbad) if (nbad ne 0) then rv[bad] = !values.f_nan ; window,/free,xsiz=100*5,ysiz=100*3 ; tvscl,congrid(alog10(rotate(reform(tp[1,*,200:239]>1),-5)),100,100),0 ; s = [1,74,93,33.,92,28,37,26,48,70,32,50,51,144,48] ; ss = reverse(sort(s)) ; for i = 1, nbl-1 do begin ; tvscl,congrid(rotate(reform(rv(ss[i-1],*,200:239))>0<1,-5),100,100),i ; xyouts,2+(i mod 5)*100,(2-i/5)*100+20,bl_str[ss[i-1]],color=255,/dev ; endfor return,rv end