-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathmpgaussfit.ct
185 lines (148 loc) · 7.27 KB
/
mpgaussfit.ct
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
;*******************************************************************************
pro mpgaussfit, wavescale, flux, err, continuum, cerr, lineinfo, $
linedata = linedata, fit = fit, noplot = noplot
;*******************************************************************************
;*******************************************************************************
; Compute equivalent width
;*******************************************************************************
continsub = flux - continuum
normflux = flux / continuum
nerr = sqrt((err / flux)^2 + (cerr / continuum)^2) * normflux
eqwidth=tsum(wavescale, 1.0 - normflux)
eqwidth_err=sqrt(tsum(wavescale, nerr^2))
total_err = sqrt(err^2. + cerr^2.)
;*******************************************************************************
; Define fit limits & starting coefs for emission & absorption lines
;*******************************************************************************
parinfo = replicate({value: 0.0D, fixed: 0, limited: [1,1], $
limits:[0.0D, 0.0D], tied: ""}, 6)
parinfo(0).value = abs(max(continsub)) ; starting ampl
parinfo(0).limits(0) = 0 ; ampl min
parinfo(0).limits(1) = abs(max(continsub))*1.6 + 0.1 ; ampl max
parinfo(1).value = lineinfo.centr ; starting line center
parinfo(1).limits(0) = lineinfo.centr - 3.0 ; line center min
parinfo(1).limits(1) = lineinfo.centr + 3.0 ; line center max
parinfo(2).value = 2.5 ; starting fwhm
parinfo(2).limits(0) = 1.0 ; fwhm min
parinfo(2).limits(1) = 15.0 ; fwhm max
parinfo(3).value = -1.0*abs(min(continsub)) ; starting ampl
parinfo(3).limits(0) = abs(min(continsub))*(-1.6)-0.1 ; ampl min
parinfo(3).limits(1) = 0 ; ampl max
parinfo(4).value = lineinfo.centr ; starting line center
parinfo(4).limits(0) = lineinfo.centr - 8.0 ; line center min
parinfo(4).limits(1) = lineinfo.centr + 8.0 ; line center max
parinfo(5).value = 10.0 ; starting fwhm
parinfo(5).limits(0) = 3.0 ; fwhm min
parinfo(5).limits(1) = 30.0 ; fwhm max
if (lineinfo.type eq 'em') then begin
parinfo([3,4,5]).limits = 0
parinfo([3,4,5]).value = 0
parinfo([3,4,5]).fixed = 1
endif
if (lineinfo.type eq 'abs') then begin
parinfo([0,1,2]).limits = 0
parinfo([0,1,2]).value = 0
parinfo([0,1,2]).fixed = 1
endif
;*******************************************************************************
; Perform fit on spectrum
;*******************************************************************************
wavescale = wavescale * 1.D
continsub = continsub * 1.D
total_err = total_err * 1.D
gcoef = parinfo(*).value * 1.D
gcoeflast = gcoef * 0
itera = 0
while total(gcoef - gcoeflast) ne 0.0 do begin
gcoeflast = gcoef
gfit = mpcurvefit(wavescale, continsub, 1.0/total_err^2, gcoef, sigma, $
function_name = 'doublegauss', parinfo = parinfo, errmsg = errmsg, $
iter = niter, status = status, covar = covar, chisq = chisq, $
autoderivative = 0, /quiet)
parinfo.value = gcoef
itera = itera + niter
if itera gt 300 then gcoeflast = gcoef
endwhile
;*******************************************************************************
; Compute errors on fitted parameters the curvefit way
;*******************************************************************************
weights = 1.0 / total_err^2
doublegauss, wavescale, gcoef, y, pder
nterms = n_elements(gcoef) ; # of parameters
diag = lindgen(nterms)*(nterms+1) ; Subscripts of diagonal elements
alpha = transpose(pder) # (Weights # (fltarr(nterms)+1)*pder)
sigma = sqrt( 1.0 / alpha[diag] )
sigma_bad=where(finite(sigma) EQ 0,cnt) ;clean up the zero pder
IF cnt GE 1 THEN sigma[sigma_bad]=0
;*******************************************************************************
; Compute the S/N of the fit
;*******************************************************************************
gausspars, gcoef[0:2], sigma[0:2], struct = em
gausspars, gcoef[3:5], sigma[3:5], struct = abs
em_sigma = abs(em.flux / em.flux_err)
abs_sigma = abs(abs.flux / abs.flux_err)
if lineinfo.type eq 'em' then nsigma = em_sigma
if lineinfo.type eq 'abs' then nsigma = abs_sigma
if lineinfo.type eq 'both' then begin
nsigma = (em.flux - abs.flux) / sqrt(em.flux_err^2 + abs.flux_err^2)
if em.flux eq 0 then nsigma = abs_sigma
if abs.flux eq 0 then nsigma = em_sigma
endif
;*******************************************************************************
; If not well fit then set coef to zero
;*******************************************************************************
if nsigma le 1.0 or not finite(nsigma) then begin
gcoef[*] = 0.0
sigma[*] = 0.0
fit = gfit * 0
nsigma=0.
endif else fit = gfit
gausspars, gcoef[0:2], sigma[0:2], struct = em
gausspars, gcoef[3:5], sigma[3:5], struct = abs
if lineinfo.type eq 'em' or lineinfo.type eq 'both' then g = em else g = abs
;*******************************************************************************
; Overplot fit and output nsigma
;*******************************************************************************
if not keyword_set(noplot) and nsigma gt 1.0 then begin
if nsigma gt 1.0 then fitcolor = !blue
if nsigma ge 1.96 then fitcolor = !dgreen
if nsigma ge 2.33 then fitcolor = !red
oplot, wavescale - lineinfo.centr, fit + continuum, color=fitcolor, thick=2
xyouts, -33, !y.crange[1]+3*(!y.crange[1]-!y.crange[0])/n_elements(!y.tickv),$
strn(nsigma,format='(f5.2)')+'!4r!x',charsize=1.0
endif
;*******************************************************************************
; Store results of fit
;*******************************************************************************
tiny=1E-37 ; to reduce errors
linedata.eqwidth = eqwidth
linedata.eqwidth_err = eqwidth_err
cflux = continuum[value_to_index(wavescale, g.center)]
linedata.continuum = cflux
linedata.cerr = cerr
for i = 0, 5 do linedata.(i + 4) = g.(i)
linedata.geqwidth = -1 * g.flux / (cflux + tiny)
linedata.geqwidth_err = g.flux_err / (cflux + tiny)
linedata.sn = nsigma
if lineinfo.type eq 'both' then begin
linedata = [linedata, linedata]
cflux = continuum[value_to_index(wavescale, abs.center)]
linedata[1].continuum = cflux
for i = 0, 5 do linedata[1].(i + 4) = abs.(i)
linedata[1].geqwidth = -1 * abs.flux / (cflux + tiny)
linedata[1].geqwidth_err = abs.flux_err / (cflux + tiny)
endif
;*******************************************************************************
; Print results of fit
;*******************************************************************************
print, string(format = '(A11)', lineinfo[0].line) + $
string(format = '(F9.2)', linedata[0].eqwidth ) + $
string(format = '(F8.2)', linedata[0].eqwidth_err) + $
string(format = '(F9.2)', linedata[0].geqwidth) + $
string(format = '(F9.2)', linedata[0].flux) + $
string(format = '(F10.2)', linedata[0].flux_err) + $
string(format = '(F8.2)', linedata[0].fwhm) + $
string(format = '(F7.1)', nsigma) + $
string(format = '(I6)', itera)
;*******************************************************************************
end