-
Notifications
You must be signed in to change notification settings - Fork 0
/
cairs_gauss.py
312 lines (261 loc) · 10.2 KB
/
cairs_gauss.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
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
import numpy
import sys
# sys.path.remove('/usr/local/lib/python2.7/dist-packages/cubature-0.13.0_dev-py2.7-linux-x86_64.egg')
#print sys.path
#sys.path.insert(1,
# 'home/matt/apt/envVirtual/lib/python2.7/site-packages/cubature/dist/cubature-0.13.0_dev-py2.7-linux-x86_64.egg')
#sys.path.insert(1, '/home/matt/apt/envVirtual/lib/python2.7/site-packages/cubature')
from cubature import cubature
from math import sqrt, exp
from cairs_coor import Coor, Domain
# from datetime import timedelta
#import pdb
import logging
logger = logging.getLogger('cairs')
class Mean(object):
def __init__(self, mean=0):
self.mean = mean
print('Prior has a constant mean of %s' % mean)
self.extend_v = None
self.pos_v = None
self.d = None
self.counter = 0
#st=StaticContext()
#self.REF_TIME=st.REF_TIME
def f_mu(self, d):
"""
#vector of domain extend
mean function
:param d: domain
:type d: Domain
"""
if type(d) is Domain:
self.d = d
self.extend_v = numpy.array([d.extend.x,
d.extend.y,
d.extend.time], dtype=numpy.float64)
# vector of positions
self.pos_v = numpy.array([d.position.x,
d.position.y,
d.position.time], dtype=numpy.float64)
lower = []
for p, v in zip(self.pos_v, self.extend_v):
if v != 0:
lower.append(p)
#lower=min(lower)
upper = []
for p, v in zip(self.pos_v, self.extend_v):
if v != 0:
if v + p != 0:
x = p + v
upper.append(x)
upper = numpy.array(upper, dtype=numpy.float64)
lower = numpy.array(lower, dtype=numpy.float64)
#I = integrate.fixed_quad(self.f_int,a=lower[0],b=upper[0])
I = cubature(self.f_int, ndim=lower.shape[0],
fdim=upper.shape[0], maxEval=10000,
xmin=lower, xmax=upper, adaptive='p',
)
return I[0][0]
else:
return self.mean
def f_int(self, x_array):
"""
:brief : construct function to integrate over
:param v: numbers
:type v: list
:return:
:rtype:
"""
kk = 0
for i in range(0, 2, 1):
if self.extend_v[i] != 0:
self.pos_v[i] = x_array[kk]
kk += 1
coorTMP = Coor(self.pos_v[0], self.pos_v[1], self.pos_v[2])
coorTMP = Coor.rotate(coorTMP, self.d.position, self.d.angle)
return self.f_mu(coorTMP)
class Covarience(object):
def __init__(self, sigma=10, cov_o=None, l_spatial=3000, l_temporal=60, gamma=1):
self.sigma = sigma
self.cov_o = cov_o
self.extend_v = None
self.pos_v = None
self.d1 = None
self.d2 = None
self.l_spatial = l_spatial
self.l_temporal = l_temporal
self.gamma = gamma
if gamma < 0 or gamma > 2:
sys.exit("Gamma must be in [0,2]")
def f_cov(self, d1, d2):
"""
:param d1:
:type d1:
:param d2:
:type d2:
:return:
:rtype:
"""
#logger.debug("d1 = %s"%d1)
#logger.debug("d2 = %s"%d2)
mark = False
if type(d1) is Coor and type(d2) is Coor:
#logger.debug("self.c1 %s" % d1)
#logger.debug("self.c2 %s" % d2)
var = self.sigma ** 2
dist_spatial = sqrt((d1.x - d2.x) ** 2 + (d1.y - d2.y) ** 2)
dist_temporal = abs(d1.time - d2.time)
timeDiv = dist_temporal / self.l_temporal
expFeed = (-1) * ((dist_spatial / self.l_spatial) ** self.gamma) - (timeDiv ** self.gamma)
cov = var * exp(expFeed)
#logger.debug('134')
#logger.debug("f_cov %s" % cov)
return cov
else:
# -----------
## covariance (kernel) function for ( Domain, Domain )
if type(d1) is Coor and type(d2) is Domain: ## --- for ( Coor, Domain )
d1 = Domain(d1, Coor(0, 0), 0)
mark = True
if type(d2) is Coor and type(d1) is Domain and not mark:
d2 = Domain(d2, Coor(0, 0), 0)
self.extend_v = numpy.array([d1.extend.x,
d1.extend.y,
d1.extend.time,
d2.extend.x,
d2.extend.y,
d2.extend.time], dtype=numpy.float64)
# # vector of positions
self.pos_v = numpy.array([d1.position.x,
d1.position.y,
d1.position.time,
d2.position.x,
d2.position.y,
d2.position.time], dtype=numpy.float64)
self.d1 = d1
self.d2 = d2
# # construct function to integrate over
lower = list()
for p, v in zip(self.pos_v, self.extend_v):
if v != 0:
lower.append(p)
upper = list()
for p, v in zip(self.pos_v, self.extend_v):
if v != 0 and v + p != 0:
x = p + v
upper.append(x)
## integrate
#sumExtend_v=sum(self.extend_v)
#TODOif sumExtend_v != 0 and sumExtend_v >2: # for >2-dimensional integration
dim = 0
for ex in self.extend_v:
if ex != 0:
dim += 1
upper = numpy.array(upper, dtype=numpy.float64)
lower = numpy.array(lower, dtype=numpy.float64)
if dim > 2:
#logger.debug("dim >2")
# "I = cubature(self.f_int, xmin=lower,xmax=upper,ndim=lower.shape[0],fdim=upper.shape[0], maxEval=10000)")
#I = integrate.nquad(self.f_int,[lower,upper])
I = cubature(self.f_int, xmin=lower, xmax=upper, ndim=lower.shape[0], fdim=upper.shape[0],
maxEval=10000)
#I = integrate.fixed_quad(self.f_int,a=lower[0],b=upper[0])
else:
#logger.debug("dim <= 2")
#logger.debug("self.d1 %s" % self.d1)
#logger.debug("self.d2 %s" % self.d2)
#logger.debug("upper = %s" % upper)
#logger.debug("lower = %s" % lower)
#logger.debug('199')
#if len(upper)>2:
#I = integrate.quad(self.f_int,a=lower[0],b=upper[0])
#logger.debug('upper.shape %s' % upper.shape)
#.debug('upper.shape[0] %s' % upper.shape[0])
I = cubature(self.f_int, ndim=lower.shape[0],
fdim=upper.shape[0], maxEval=10000,
xmin=lower, xmax=upper, adaptive='p')
#logger.error(logger.findCaller())
#logger.debug('212 cubature --------')
#logger.debug("I[0] = %s" % I[0])
try:
return I[0][0]
except:
return I[0]
logger.error('f_cov nothing happened')
def f_int(self, x_array):
#logger.debug('f_int')
# # change the coordiates for integration
kk = 0
#logger.debug('x_array %s' % numpy.asarray(x_array))
for i in range(0, 5, 1):
if self.extend_v[i] != 0:
if type(x_array) in [float, numpy.float64]:
#logger.error("x_array = %s" % x_array)
self.pos_v[i] = x_array
else:
self.pos_v[i] = x_array[kk]
#break
kk += 1
#logger.debug("len(self.pos_v) = %s"%len(self.pos_v))
#logger.error(logger.findCaller())
x = Coor.rotate(Coor(self.pos_v[0], self.pos_v[1], self.pos_v[2]),
self.d1.position,
self.d1.angle)
#logger.error('x = Coor.rotate')
y = Coor.rotate(Coor(self.pos_v[3], self.pos_v[4], self.pos_v[5]),
self.d2.position,
self.d2.angle)
#logger.error(logger.findCaller())
#logger.debug('242')
cov = self.f_cov(x, y)
r = list()
if len(numpy.asarray(x_array)) == 1:
return cov
else:
for i in range(0, len(numpy.asarray(x_array)), 1):
r.append(cov)
return r
def make_cov(self, loc_1, loc_2):
"""
:param loc_1:
:type loc_1:
:param loc_2:
:type loc_2:
:return:
:rtype:
"""
if loc_1 != loc_2:
sigma = numpy.zeros(shape=(len(loc_1), len(loc_2)), dtype=numpy.float64)
for i,l1 in enumerate(loc_1):
for j,l2 in enumerate(loc_2):
sigma[i,j]=(self.f_cov(l1, l2)) # TODO check
#logger.debug("sigma makexcov = %s"%sigma)
else:
n = len(loc_1)
sigma = numpy.zeros(shape=(n, n), dtype=numpy.float64)
for j in range(0, n, 1):
for i in range(j, n, 1):
if i != j:
x = self.f_cov(loc_1[i], loc_2[j])
sigma[i, j] = sigma[j, i] = x
else:
x = self.f_cov(loc_1[i], loc_2[j])
sigma[i, j] = x
return sigma
@staticmethod
def log_p_prior(locations, samp_dict, i_sample, mu, sigma):
R = list()
for loc in locations:
a = samp_dict[loc]
R.append(a[i_sample])
D = list()
for i, m in enumerate(mu):
a = R[i] - m
D.append(a)
#logger.debug("R = %s" % R)
#logger.debug("D = %s" % D)
# compute D' inv(Sigma) * D
neg_log_p = numpy.dot(numpy.transpose(D), numpy.linalg.inv(sigma))
neg_log_p = numpy.dot(neg_log_p, D) #SOLVED
return -neg_log_p