-
Notifications
You must be signed in to change notification settings - Fork 0
/
spec_fit.py
183 lines (162 loc) · 6.32 KB
/
spec_fit.py
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
#Just need for spectra classes
import corrspline as corr
#For rebinning of models
import model_bin as mb
import ipdb
import numpy as np
import matplotlib.pyplot as plt
import dill
import scipy
from scipy import optimize
def cfit_model(p,full_model,wavelens):
#S make a chunk of the model shifted by doppler factor z
ipmodel = mb.numconv(full_model.data,gaussian(p))
factor = np.polyval(p[:1:-1],full_model.wavelens)
unbin_model = factor*ipmodel/np.max(ipmodel)
# model = mb.john_rebin(full_model.wavelens,full_model.data,wavelens,p[0])
model = mb.john_rebin(full_model.wavelens,unbin_model,wavelens,p[0])
return model
#S continuum subtraction
# coeffs = np.polyfit(wavelens,model,5)
#S do we need to add 1? will just come out in the wash. might want to boost
#S the observed though
# cs_model = model - np.polyval(coeffs,wavelens)
# convmodel = mb.numconv(model,gaussian(p))
factor = np.polyval(p[:1:-1],wavelens)
return factor*convmodel/np.max(convmodel)
# return factor*model/np.max(model)
def fit_model(p,full_model,wavelens):
#S make a chunk of the model shifted by doppler factor z
model = mb.john_rebin(full_model.wavelens,full_model.data,wavelens,p[0])
#S continuum subtraction
coeffs = np.polyfit(wavelens,model,5)
#S do we need to add 1? will just come out in the wash. might want to boost
#S the observed though
cs_model = model - np.polyval(coeffs,wavelens)
# return p[1]*model/np.max(model)
factor = np.polyval(p[:0:-1],wavelens)
return factor*model/np.max(model)
def gaussian(p):
xip = np.arange(-10,10+.25,.25)
return np.exp(-(xip/p[1])**2.)
def err_func(p,model,star,order):
wavelens = star.wavelens[order]
inds = star.inds[str(order)]
fitspec = star.data[order][inds]
return (cfit_model(p,model,wavelens)[inds] -\
fitspec)/np.sqrt(np.abs(fitspec))
if __name__ == '__main__':
# full_model = corr.highres_spec('./t05500_g+0.5_m10p04_hr.fits')
full_model = corr.highres_spec('./t05500_g+4.0_p00p00_hrplc.fits')
suffix = '_norm2'
blg = dill.load(open('./BLG0966'+suffix+'.pkl','rb'))
hr = dill.load(open('./HR4963'+suffix+'.pkl','rb'))
hd = dill.load(open('./HD142527'+suffix+'.pkl','rb'))
# ipdb.set_trace()
rv=[]
rver = []
"""
plt.plot(np.arange(len(blg.data)),np.sqrt(np.median(blg.data,axis=1)),'o')
plt.xlabel('Order')
plt.ylabel(r'$\sqrt{{\bf MEDIAN}(N)}$')
plt.show()
ipdb.set_trace()
"""
ipdb.set_trace()
offset = 250
params = []
blg.inds = {}
params_list=[]
# for order in [3]:
for order in np.arange(len(blg.data)-28)+2:
# sigma clip from the top
zinds = np.where(blg.data[order]>0.)[0]
inds = zinds
ct = len(inds)
ran = np.arange(len(blg.data[order]))
dct = 1
while dct>0:#oldlen != newlen:
med = np.median(blg.data[order][inds])
std = np.std(blg.data[order][inds])
inds = zinds[np.where(blg.data[order][zinds]<med+3*std)[0]]
newct = len(inds)
dct = ct-newct
ct = newct
blg.inds[str(order)] = inds
# coeffs = np.polyfit(blg.wavelens[order][ind],blg.data[order][ind],2,\
# w=1/np.sqrt(blg.data[order][ind]))
# cscorr = np.polyval(coeffs,blg.wavelens[order])
# fit_data = (blg.data[0]-cscorr+median)
p0 = [0.00014,1.,np.median(blg.data[order]),0.,0.]
# p0 = [0.00014,np.median(blg.data[order]),0.]#,0.]
out=scipy.optimize.leastsq(err_func,p0,args=\
(full_model,blg,order),\
full_output=1,maxfev=1000)
message=out[3]
ier=out[4]
print 'Fitter status:',ier,' Message: ',message
p1 = out[0]
# if out[1] == None:
# ipdb.set_trace()
# continue
jacob=out[1]
mydict=out[2]
resids=mydict['fvec']
# covar = np.std(resids)**2*jacob
chisq = np.sum(resids**2)
degs_frdm=len(blg.wavelens[order])-len(p1)
red_chisq = chisq/degs_frdm
i=0
for u in p1:
print 'Param: ', i+1, ': ',u,' +/-'#,np.sqrt(covar[i,i])
i+=1
print 'Chisq: ',chisq,' Reduced Chisq: ',red_chisq
params.append(p1)
rv.append(p1[0]*2.99e8)
# rver.append(np.sqrt(covar[0,0])*3e8)
# ipdb.set_trace()
params_list.append(p1)
inds = blg.inds[str(order)]
plt.plot(inds,offset*order+\
blg.data[order][inds],'b',zorder=1)
plt.plot(inds,offset*order+\
cfit_model(p0,full_model,blg.wavelens[order])[inds],\
'g',zorder=2)
plt.plot(inds,offset*order+\
cfit_model(p1,full_model,blg.wavelens[order])[inds],\
'r',zorder=2,linewidth=2)
"""
plt.plot(blg.wavelens[order][inds],\
blg.data[order][inds],'b',zorder=1)
# plt.plot(blg.wavelens[order][inds],\
# cfit_model(p0,full_model,blg.wavelens[order])[inds],\
# 'g',zorder=2)
plt.plot(blg.wavelens[order][inds],\
cfit_model(p1,full_model,blg.wavelens[order])[inds],\
'r',zorder=2,linewidth=2)
"""
"""
plt.plot(blg.wavelens[order],blg.data[order]+offset*order,'b',zorder=1)
plt.plot(blg.wavelens[order],offset*order+fit_model(p0,full_model,blg.wavelens[order]),\
'g',zorder=2)
plt.plot(blg.wavelens[order],offset*order+fit_model(p1,full_model,blg.wavelens[order]),\
'r',zorder=2)
"""
plt.xlabel('Wavelength')
plt.ylabel('Counts')#''Arbitrary')
# plt.axis([8400,9000,0,500])
plt.show()
barycorr = 24167.
ipdb.set_trace()
rv = (np.array(rv)+barycorr)/1000.
# rver = np.array(rver)/1000.
plt.plot(np.arange(len(rv)),rv,'o',label='Individual order RV')
trv = np.concatenate([rv[0:5],rv[7:11],rv[13:]])
all_rv = 49.791 + barycorr/1000.
plt.plot([0,32],[all_rv,all_rv],'r-',label='RV from simul. fit')
# plt.errorbar(np.arange(len(rv)),rv,yerr=rver)
plt.xlabel('"Order"')
plt.ylabel('RV [km s$^{-1}$]')
plt.legend()
plt.show()
ipdb.set_trace()