-
Notifications
You must be signed in to change notification settings - Fork 0
/
plot_fisher_real_pdf256.py
159 lines (106 loc) · 3.72 KB
/
plot_fisher_real_pdf256.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
import numpy
import matplotlib.pyplot as plt
import matplotlib.lines as mlines
from matplotlib.patches import Ellipse
N = 20.0 * 384.0
fisher = 'data_real/F_1024_test_pdf256_1.npz'
basetitle = 'PDF 256 scaling'
fisher2 = 'data_real/F_1024_test_pdf256_2.npz'
fisher3 = 'data_real/F_1024_test_pdf256_3.npz'
fisher4 = 'data_real/F_1024_test_pdf256_2_3.npz'
fisher5 = 'data_real/F_1024_test_pdf256_4.npz'
#fisher = 'data/F_8192_fullcov_7.npz'
#basetitle = 'PDF (nside 1024)'
#fisher = 'data/F_8192_fullcov_3.npz'
#basetitle = 'PDF (nside 4096)'
data = numpy.load(fisher)
F = data['F'] #* (N - 400.0 - 2.0) / (N - 2.0)
params = data['params']
data = numpy.load(fisher2)
F2 = data['F'] #* (N - 800.0 - 2.0) / (N - 2.0)
data = numpy.load(fisher3)
F3 = data['F'] #* (N - 1600.0 - 2.0) / (N - 2.0)
data = numpy.load(fisher4)
F4 = data['F'] #* (N - 2400.0 - 2.0) / (N - 2.0)
data = numpy.load(fisher5)
F5 = data['F'] #* (N - 3200.0 - 2.0) / (N - 2.0)
n = len(params)
names = params
print names
centers = {'h':0.69,'ok':0.0, 'w':-1.0, 's8':0.82, 'om':0.29, 'q':1.0}
#print F4
F[-2,-2] += 1/(0.1*0.1)
F2[-2,-2] += 1/(0.1*0.1)
F3[-2,-2] += 1/(0.1*0.1)
F4[-2,-2] += 1/(0.1*0.1)
F5[-2,-2] += 1/(0.1*0.1)
F[-1,-1] += 1/(0.05*0.05)
F2[-1,-1] += 1/(0.05*0.05)
F3[-1,-1] += 1/(0.05*0.05)
F4[-1,-1] += 1/(0.05*0.05)
F5[-1,-1] += 1/(0.05*0.05)
#F[4,4] += 1/(0.0001*0.0001)
#F2[4,4] += 1/(0.0001*0.0001)
#F3[4,4] += 1/(0.0001*0.0001)
#F4[4,4] += 1/(0.0001*0.0001)
#F5[4,4] += 1/(0.0001*0.0001)
#F[1,1] += 1/(0.0001*0.0001)
#F2[1,1] += 1/(0.0001*0.0001)
#F3[1,1] += 1/(0.0001*0.0001)
#F4[1,1] += 1/(0.0001*0.0001)
#F5[1,1] += 1/(0.0001*0.0001)
#F[0,0] += 1/(0.1*0.1)
#F2[0,0] += 1/(0.1*0.1)
#F3[0,0] += 1/(0.1*0.1)
#F4[0,0] += 1/(0.1*0.1)
#F5[0,0] += 1/(0.1*0.1)
def plot_fisher(F, basetitle, i,j, color, rescale=False):
F_inv = numpy.linalg.pinv(F)
ixgrid = numpy.ix_([j,i],[j,i])
subM = F_inv[ixgrid]
print subM
sigma1 = numpy.sqrt(subM[0,0])
sigma2 = numpy.sqrt(subM[1,1])
subF = numpy.linalg.pinv(subM)
l, v = numpy.linalg.eig(subM)
l = numpy.sqrt(l)
area = l[0]*l[1]*numpy.pi
print names[j] + ' sd: ' + str(sigma1)
print names[i] + ' sd: ' + str(sigma2)
print 'area: ' + str(area)
print 'fom: ' + str(1.0/area)
ax = plt.subplot(111)
ax.set_title(basetitle + ': ' + names[j] + ' vs ' + names[i])
ax.set_xlabel(names[j])
ax.set_ylabel(names[i])
center_i = centers[params[i]]
center_j = centers[params[j]]
ell = Ellipse(xy=(center_j,center_i),width=l[0]*2, height=l[1]*2, angle=numpy.rad2deg(-numpy.arccos(v[0,0])))
ell.set_facecolor('none')
ell.set_edgecolor(color)
ax.add_artist(ell)
if rescale:
plt.xlim([center_j-numpy.sqrt(subM[0,0])*1.2,center_j+numpy.sqrt(subM[0,0])*1.2])
plt.ylim([center_i-numpy.sqrt(subM[1,1])*1.2,center_i+numpy.sqrt(subM[1,1])*1.2])
plt.scatter([center_j],[center_i])
for i in range(n):
for j in range(i):
if names[i] != 's8':
continue
if names[j] != 'om':
continue
handles = []
plot_fisher(F, basetitle, i,j,'blue',False)
plt.plot([],color='blue',label='400 bins')
plot_fisher(F2, basetitle, i, j, 'green',True)
plt.plot([],color='green',label='800 bins')
plot_fisher(F3, basetitle, i, j, 'red', False)
plt.plot([],color='red',label='1600 bins')
plot_fisher(F4,basetitle,i,j, 'black',False)
plt.plot([],color='black',label='2400 bins')
plot_fisher(F5, basetitle, i, j, 'purple', False)
plt.plot([],color='purple',label='3200 bins')
plt.legend()
plt.xlim([0.15,0.43])
plt.ylim([0.61,1.03])
plt.show()