/
shear.py
148 lines (125 loc) · 4.64 KB
/
shear.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
"""
shear.py
Author: Marzia Rivi (2018)
arguments: -nf number of files with name ellipticities<n>.txt
Compute shear as a weighted mean of the galaxies ellipticity
use bootstrap to compute standard deviation
Use data simulations where shapes are generated to be zero on average.
Opposite ellipticities are consecutives. Therefore for each bad data,
the opposite is removed too from the shear computation.
"""
import sys
import argparse
import math
import numpy as np
#from pylab import *
import astropy.stats as astro
parser = argparse.ArgumentParser(description='bootstrap_std')
parser.add_argument('-nf',dest='nfiles', type=int, default=1, help='Number of measurement files')
args = parser.parse_args(sys.argv[1:])
e_max = 0.804 # ellipticity cutoff
e_0 = 0.0256 # circularity parameter
a = 0.2539 # dispersion
Nfactor = 2.43180252985281 # A=1/0.4112176 normalization factor
def prior_ellipticity(ee1,ee2):
emod = np.sqrt(np.multiply(ee1,ee1)+np.multiply(ee2,ee2))
p_e = Nfactor*np.divide(np.multiply(emod,(1.-np.exp((emod-e_max)/a))),np.multiply(1.+emod,np.sqrt(np.multiply(emod,emod)+e_0*e_0)))
# read ellipticities
me1=[]
me2=[]
err1=[]
err2=[]
SNR=[]
w=[]
k=0
remove = 0
nfiles=args.nfiles
bad = 0
num=0
#ngal = 10000
try:
while num<nfiles:
name = "ellipticities%d.txt"%(num)
# data=np.loadtxt(name, skiprows=1, delimiter='|',usecols=(1,2,3,4,5,6))
data=np.loadtxt(name, skiprows=1, delimiter='|',usecols=(2,3,5,6,7,8))
ngal = len(data)
print ngal
i=0
while i<len(data): # and k<ngal:
error1 = data[i,1]
error2 = data[i,3]
var = data[i,4]
SNRvalue = data[i,5]
if var > 1e-5 and error1>1e-3 and error2>1e-3 and SNRvalue >= 10:
if remove == 1:
me1[k] = data[i,0]
err1[k] = error2
me2[k] = data[i,2]
err1[k] = error1
w[k] = var*e_max*e_max/(e_max*e_max-2*var)
SNR[k] = SNRvalue
else:
me1.append(data[i,0])
err1.append(error2)
me2.append(data[i,2])
err1.append(error1)
w.append(var*e_max*e_max/(e_max*e_max-2*var))
SNR.append(SNRvalue) # 8
k = k+1
remove = 0
else: # remove the opposite too
if i%2 != 0: # odd line: opposite is the previous one
remove = 1 # remove previous element by overlapping it with the following
k = k-1
else: # even line: opposite is the next one
i = i+1 # skip next data line
print "bad measure:",data[i,:]
bad = bad + 1
i=i+1
num = num + 1
except:
print 'ERROR!'
# if last line is bad measure
if remove == 1:
k = k-1
me1 = np.delete(me1,k)
me2 = np.delete(me2,k)
SNR = np.delete(SNR,k)
w = np.delete(w,k)
print "mean SNR ",np.mean(SNR), "median SNR ",np.median(SNR)
print "ngal: ", len(me1)," bad: ",bad
print "min SNR",np.min(SNR)
print 'measured mean: ',np.mean(me1), np.mean(me2)
# compute shape noise
pe = prior_ellipticity(me1,me2)
mean1 = np.average(me1,weights=pe)
mean2 = np.average(me2,weights=pe)
mean12 = np.average(np.multiply(me1,me2),weights=pe)
var12 = mean12 - mean1*mean2
mean11 = np.average(np.multiply(me1,me1),weights=pe)
mean22 = np.average(np.multiply(me2,me2),weights=pe)
var1 = mean11 - mean1*mean1
var2 = mean22 - mean2*mean2
sigma_shape = np.sqrt(var1*var2-var12*var12)
print "shape noise: ",sigma_shape
# compute shear
norient = 2
ngal = len(me1)/norient
print norient*ngal
w = np.divide(1.,w+sigma_shape)
data = np.array(me1).reshape(ngal,norient)
weight = np.array(w).reshape(ngal,norient) #np.ones((ngal,norient))
alldata=np.concatenate((data,weight),axis=1)
data = np.array(me2).reshape(ngal,norient)
weight = np.array(w).reshape(ngal,norient)
# concatenate all and reshape as 4-dim array of dimensions (ngal,ncomponents,ntypes,norient), where types = measure, weight
alldata=np.concatenate((alldata,data,weight),axis=1).reshape(ngal,2,2,norient)
bootnum = 1000
bootsamples = astro.bootstrap(alldata,bootnum)
e1_shears=np.zeros(bootnum)
e2_shears=np.zeros(bootnum)
for i in range(bootnum):
e1_shears[i] = np.average(bootsamples[i,:,0,0,:],weights=bootsamples[i,:,0,1,:])
e2_shears[i] = np.average(bootsamples[i,:,1,0,:],weights=bootsamples[i,:,1,1,:])
print 'e1_shear: mean = ',np.mean(e1_shears),' std = ',np.std(e1_shears)
print 'e2_shear: mean = ',np.mean(e2_shears),' std = ',np.std(e2_shears)