-
Notifications
You must be signed in to change notification settings - Fork 0
/
cst.py
195 lines (140 loc) · 6.29 KB
/
cst.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
184
185
186
187
188
189
190
191
192
193
import sys
utility_folder = '../utility/'
sys.path.insert(0, utility_folder)
import csv
from math import *
from matplotlib.pylab import *
from sys import argv
import util as ut
import preproc_prof as pp
import scipy.optimize
#class Profile:
class Cst:
def __init__(self, a):
self._N = a[0:2]
self._A = a[2:len(a)]
def __call__(self, x):
n = len(self._A)
S = 0
for i in xrange(n):
S += self._A[i]*self._bin_coef(i,n-1)*(1-x)**(n-1-i)*x**i
return self._class_function(x)*S
def _class_function(self, x):
return x**self._N[0]*(1-x)**self._N[1]
def _bin_coef(self, i, n):
return factorial(n)/(factorial(i)*factorial(n-i))
class BF_cst:
def __init__(self, xy_p, c_err = 1.0, n_ord = 2.0):
self._xy_p = xy_p
self._c_err = c_err
self._n_ord = n_ord
def __call__(self, A):
self._cst = Cst(A)
return np.linalg.norm(self._err(), ord = self._n_ord)
def _err(self):
return np.where(self._xy_p[1,:] != 0.0, abs(self._xy_p[1,:] - self._cst(self._xy_p[0,:]))/abs(self._xy_p[1,:])**self._c_err, abs(self._xy_p[1,:] - self._cst(self._xy_p[0,:])))
@staticmethod
def init_cfc( cfc_ig1 = 0.4, cfc_ig2 = 0.4):
N = np.ones(2)
N[0] *= cfc_ig1
N[1] *= cfc_ig2
return N
@staticmethod
def init_A(poly_gr = 5, A_ig = 0.8):
A = np.ones(poly_gr+1)
A *= A_ig
return A
@staticmethod
def bound_cfc( cfc_lb1 = 0.1, cfc_ub1 = 1.5, cfc_lb2 = 0.1 , cfc_ub2 = 1.4):
N = np.ones((2,2), dtype = float)
N[0,0] *= cfc_lb1
N[1,0] *= cfc_lb2
N[0,1] *= cfc_ub1
N[1,1] *= cfc_ub2
return N
@staticmethod
def bound_A(poly_gr = 4, A_lb = -1.5, A_ub = +1.5):
A = np.ones((poly_gr + 1, 2), dtype = float)
A[:,0] *= A_lb
A[:,1] *= A_ub
return A
if __name__=='__main__':
script, input_file = argv
import ConfigParser
config = ConfigParser.ConfigParser()
config.read(input_file)
filename = config.get("general", "filename")
xy_ss, xy_ps = pp.import_prof(filename)
poly_gr_ss = int(config.get("ss", "poly_gr_ss"))
cfc_lb1_ss = float(config.get("ss", "cfc_lb1"))
cfc_ub1_ss = float(config.get("ss", "cfc_ub1"))
cfc_lb2_ss = float(config.get("ss", "cfc_lb2"))
cfc_ub2_ss = float(config.get("ss", "cfc_ub2"))
cfc_ig1_ss = float(config.get("ss", "cfc_ig1"))
cfc_ig2_ss = float(config.get("ss", "cfc_ig2"))
A_lb_ss = float(config.get("ss", "A_lb"))
A_ub_ss = float(config.get("ss", "A_ub"))
A_ig_ss = float(config.get("ss", "A_ig"))
c_ss = float(config.get("ss", "yp_c"))
poly_gr_ps = int(config.get("ps", "poly_gr_ps"))
cfc_lb1_ps = float(config.get("ps", "cfc_lb1"))
cfc_ub1_ps = float(config.get("ps", "cfc_ub1"))
cfc_lb2_ps = float(config.get("ps", "cfc_lb2"))
cfc_ub2_ps = float(config.get("ps", "cfc_ub2"))
cfc_ig1_ps = float(config.get("ps", "cfc_ig1"))
cfc_ig2_ps = float(config.get("ps", "cfc_ig2"))
A_lb_ps = float(config.get("ps", "A_lb"))
A_ub_ps = float(config.get("ps", "A_ub"))
A_ig_ps = float(config.get("ps", "A_ig"))
c_ps = float(config.get("ps", "yp_c"))
N_ss = BF_cst.init_cfc( cfc_ig1 = cfc_ig1_ss, cfc_ig2 = cfc_ig2_ss)
N_ps = BF_cst.init_cfc( cfc_ig1 = cfc_ig1_ps, cfc_ig2 = cfc_ig2_ps)
A_ss = BF_cst.init_A(poly_gr = poly_gr_ss, A_ig = A_ig_ss)
A_ps = BF_cst.init_A(poly_gr = poly_gr_ps, A_ig = A_ig_ps)
Ass_o = np.concatenate((N_ss, A_ss))
Aps_o = np.concatenate((N_ps, A_ps))
Nss_b = BF_cst.bound_cfc( cfc_lb1 = cfc_lb1_ss, cfc_ub1 = cfc_ub1_ss, cfc_lb2 = cfc_lb2_ss, cfc_ub2 = cfc_ub2_ss)
Nps_b = BF_cst.bound_cfc( cfc_lb1 = cfc_lb1_ps, cfc_ub1 = cfc_ub1_ps, cfc_lb2 = cfc_lb2_ps, cfc_ub2 = cfc_ub2_ps)
Ass_b = BF_cst.bound_A(poly_gr = poly_gr_ss, A_lb = A_lb_ss, A_ub = A_ub_ss)
Aps_b = BF_cst.bound_A(poly_gr = poly_gr_ps, A_lb = A_lb_ps, A_ub = A_ub_ps)
Ass_b = np.vstack((Nss_b, Ass_b))
Aps_b = np.vstack((Nps_b, Aps_b))
of_s = BF_cst(xy_ss, c_err = c_ss)
of_p = BF_cst(xy_ps, c_err = c_ps)
# Ass, nfeval_ss, rc_ss = scipy.optimize.fmin_tnc(of_ss, Ass_o, approx_grad=True, bounds = Ass_b, maxfun = 2000)
# Aps, nfeval_ps, rc_ps = scipy.optimize.fmin_tnc(of_ps, Aps_o, approx_grad=True, bounds = Aps_b, maxfun = 2000)
Ass = scipy.optimize.fmin_slsqp(of_s, Ass_o, bounds = Ass_b, iter = 1000)
Aps = scipy.optimize.fmin_slsqp(of_p, Aps_o, bounds = Aps_b, iter = 1000)
# print scipy.optimize.anneal(of_ss, Ass_o, maxeval = 2000)
# print Ass, Aps
# Ass = np.array([0.5505, 0.3403, 0.6500, 0.5169, 0.1713, 0.2214, 0.0375, 0.0397, 0.0108, 0.0173])
# Aps = np.array([0.4612, 0.2358, -0.4779, -1.0122, 0.1103, -0.0936, -0.0266, 0.0121, 0.0008, - 0.0101])
print Aps, Ass
# cst_ps = cst(xy_ps[0,:], Aps)
# cst_ss = cst(xy_ss[0,:], Ass)
cst_ps = Cst(Aps)
cst_ss = Cst(Ass)
# print of_ss(Ass), of_ps(Aps)
f, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, sharex= True, sharey = True)
ax1.plot(xy_ps[0,:], xy_ps[1,:] , 'r-', xy_ss[0,:],xy_ss[1,:], 'r-')
ax1.plot(xy_ss[0,:], cst_ss(xy_ss[0,:]), 'b-', xy_ps[0,:], cst_ps(xy_ps[0,:]), 'b-')
# ax1.plot(xy_ss[0,:], cst_ss(xy_ss[0,:]), 'b-', xy_ps[0,:], cst_ps(xy_ps[0,:]), 'b-')
ax1.axis('equal')
ax1.set_xlabel('x')
ax1.set_ylabel('y')
# ax1.set_legend(['real profile'], ['paramatrized profile'])
p21, p22, p23, p24, = ax2.plot(xy_ps[0,:], xy_ps[1,:], 'r-', xy_ps[0,:], xy_ps[1,:], 'bo', xy_ss[0,:],xy_ss[1,:],'r-', xy_ss[0,:], xy_ss[1,:], 'bo')
p25, = ax2.plot(xy_ss[0,:], cst_ss(xy_ss[0,:]), 'b-')
p26, = ax2.plot(xy_ps[0,:], cst_ps(xy_ps[0,:]), 'b-')
legend([p21, p24, p26], ["real", "real points", "parametrized"], bbox_to_anchor=(0., 1.02, 1., .102), loc=3, ncol=3, mode="expand", borderaxespad=0.)
# plt.plot(xy_ps[0,:], xy_ps[1,:], 'bo')
# plt.plot(xy_ss[0,:], xy_ss[1,:])
# plt.plot(xy_ss[0,:], xy_ss[1,:], 'bo')
ax3.plot(xy_ps[0,:], xy_ps[1,:], 'r-')
ax3.plot(xy_ss[0,:], xy_ss[1,:], 'r-')
ax4.plot(xy_ss[0,:], cst_ss(xy_ss[0,:]), 'b-')
ax4.plot(xy_ps[0,:], cst_ps(xy_ps[0,:]), 'b-')
plt.show()
raw_input('\>')
#plot(x,y)
#show()