-
Notifications
You must be signed in to change notification settings - Fork 0
/
gridworld.py
319 lines (283 loc) · 11.3 KB
/
gridworld.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
310
311
312
313
314
315
316
317
318
319
import numpy as np
''' Gridworld simulation:
| 20 | 21 | 22 | 23 | 24 |
| 15 | 16 | 17 | 18 | 19 |
| 10 | 11 | 12 | 13 | 14 |
| 5 | 6 | 7 | 8 | 9 |
| 0 | 1 | 2 | 3 | 4 |
- true position: x in [0, 5] x [0, 5]
- abstracted position: s in {0, 1, ..., 24}
- true input: u in [-pi, pi]
Grid centers:
- 0: (0.5, 0.5)
- 1: (1.5, 0.5)
- 2: (2.5, 0.5)
- 3: (3.5, 0.5)
- 4: (4.5, 0.5)
- 5: (0.5, 1.5)
- 6: (1.5, 1.5)
- 7: (2.5, 1.5)
- 8: (3.5, 1.5)
- 9: (4.5, 1.5)
- 10: (0.5, 2.5)
- 11: (1.5, 2.5)
- 12: (2.5, 2.5)
- 13: (3.5, 2.5)
- 14: (4.5, 2.5)
- 15: (0.5, 3.5)
- 16: (1.5, 3.5)
- 17: (2.5, 3.5)
- 18: (3.5, 3.5)
- 19: (4.5, .5)
- 20: (0.5, 4.5)
- 21: (1.5, 4.5)
- 22: (2.5, 4.5)
- 23: (3.5, 4.5)
- 24: (4.5, 4.5)
'''
def iterative_average(x,n,ave):
return ave+(x-ave)/(n+1)
class Grid:
def __init__(self,taglist,x_range=1000.0,y_range=1000.0):
self.taglist=taglist
self.x_range = x_range
self.y_range = y_range
self.x_bins = np.array([0., 1., 2., 3., 4.])*self.x_range/5.0
self.y_bins = np.array([0., 1., 2., 3., 4.])*self.y_range/5.0
# specifies tag density for each cell (hotspot probabilities?)
self.p = np.array([0.1, 0.1, 0.4, 0.6, 0.2,
0.1, 0.3, 0.5, 0.8, 0.3,
0.2, 0.4, 0.2, 0.2, 0.1,
0.5, 0.9, 0.4, 0.2, 0.1,
0.3, 0.6, 0.3, 0.1, 0.1])
self.tag_count = np.zeros_like(self.p)
def state(self,Agent):
pos=Agent.getPos()
return self.getAbstractPos(pos[0],pos[1])
def saveTagList(self,fname="tags"):
f=open(fname+".csv",'w')
for tag in self.taglist:
last_ping=tag.last_ping
posx=tag.pos[0]
posy=tag.pos[1]
posz=tag.pos[2]
delay=tag.delay
ID=tag.ID
bin=tag.bin
f.write("{0},{1},{2},{3},{4},{5},{6}\n".format(last_ping,posx,posy,posz,delay,ID,bin))
f.close()
def loadTagList(self,fname="tags"):
f=open(fname+".csv",'r')
taglist=[]
from AcousticTag import AcousticTag
for line in f.readlines():
last_ping,posx,posy,posz,delay,ID,bin=line.split(',')
tag = AcousticTag(ID,last_ping=float(last_ping),ping_delay=float(delay))
tag.pos=np.array([float(posx),float(posy),float(posz)])
tag.bin=bin
taglist.append(tag)
f.close()
self.taglist=taglist
return taglist
def act(self,agent, u,dt):
bins = [0., 1., 2., 3., 4.]
# simulate forward with inputs
p = 0*u.copy()
for t in range(len(u)):
# f0 dynamics
if agent.m == 0:
self.x += self.flow(self.x) +agent.m1_step(u[t],dt)
# f1 dynamics
elif agent.m == 1:
self.x += self.flow(self.x) + agent.m2_step(u[t],dt)
# randomly switch modes
if np.random.rand() < agent.m_p:
agent.m = 1
def getAbstractPos(self,x,y):
idx = np.digitize(x, self.x_bins)
idy = np.digitize(y, self.y_bins)
return 5*(idy-1) + (idx)
def abstractPosToRange(self,s):
return (s%len(self.x_bins))*self.x_range*1.0/len(self.x_bins),((s-s%len(self.x_bins))*1.0/len(self.x_bins))%len(self.y_bins)*self.y_range/len(self.y_bins),20
def getCellXY(self,x,y):
idx = np.digitize(x, self.x_bins)
idy = np.digitize(y, self.y_bins)
return idx,idy
def groundTruth(self,tau,fixed=False,time_step=1):
bin_rates=np.zeros_like(self.p)
bin_counts=0
bin_lens=np.zeros_like(self.p)
bin_sets=[set()]*len(bin_rates)
N = len(self.taglist)
time_step=min(tau/2.0,time_step)
t=0
last_meas=0
max_delay=0
for i in range(N):
if self.taglist[i].delay > max_delay:
max_delay=self.taglist[i].delay
if fixed:
self.taglist[i].last_ping=0
else:
self.taglist[i].last_ping=np.random.randn()
if time_step > max_delay:
print('time step longer than max ping rate. results may be in accurate')
while t<=200*tau or t<max_delay*10:
for tag in self.taglist:
if tag.pinging(t):
if tag.ID not in bin_sets[tag.bin]:
bin_sets[tag.bin].add(tag.ID)
bin_lens[tag.bin]+=1
tag.updatePing(t)
if last_meas+tau<=t:
rate_meas = bin_lens/tau
bin_counts+=1
bin_rates= iterative_average(rate_meas,bin_counts,bin_rates) #iteratively average rate measurement
last_meas=t
bin_sets=[set()]*len(bin_rates)
bin_lens=0*bin_lens
t+=time_step
bin_rates.shape=(5,5)
return np.flip(bin_rates,0)
def approximateField(self,tau,fixed=False,time_step=1,spacing=None,sensorRange=1,dim=2,get_points=False):
if type(spacing) == type(None): #spacinging ( xunits, yunits)
return self.groundTruth(tau,fixed,time_step)
from AcousticReciever import AcousticReciever
nx_bins = int(self.x_range/spacing[0])
ny_bins = int(self.y_range/spacing[1])
bin_rates=np.zeros((nx_bins,ny_bins))
x_bins=np.array(range(nx_bins))*self.x_range/nx_bins
y_bins=np.array(range(ny_bins))*self.y_range/ny_bins
bin_counts=0
bin_lens=np.zeros_like(bin_rates)
bin_sets=[set() for i in range((nx_bins*ny_bins))]
N = len(self.taglist)
time_step=min(tau/2.0,time_step)
#create acoustic sensor for each bin with center as its location
points = [(idx+spacing[0]/2.0,idy+spacing[1]/2.0) for idx in x_bins for idy in y_bins]
points = np.array(points)
#print(points)
#get tags in range of each sensor
inrange=[set() for i in range((nx_bins*ny_bins))]
for tag in self.taglist:
diff=points-tag.pos[:2]
dets=np.sqrt(diff[:,0]*diff[:,0]+diff[:,1]*diff[:,1])<sensorRange
dets.shape=(len(dets),1)
#print(dets,points,np.sqrt(diff[:,0]*diff[:,0]+diff[:,1]*diff[:,1]))
#print(tag.pos,tag.ID)
#input()
pos=points[np.where(dets==1)[0],:]
for px,py in pos:
idx = np.digitize(px, x_bins)
idy = np.digitize(py, y_bins)
inrange[nx_bins*(idy-1) + (idx)-1].add(tag.ID)
#print(inrange[nx_bins*(idy-1) + (idx)-1],px,py,nx_bins*(idy-1) + (idx)-1)
#print(inrange[0],inrange[1],inrange[2],inrange[3])
t=0
last_meas=0
max_delay=0
for i in range(N):
if self.taglist[i].delay > max_delay:
max_delay=self.taglist[i].delay
if fixed:
self.taglist[i].last_ping=0
else:
self.taglist[i].last_ping=np.random.randn()
if time_step > max_delay:
print('time step longer than max ping rate. results may be in accurate')
while t<=200*tau or t<max_delay*10:
for tag in self.taglist:
if tag.pinging(t):
for i in range(len(inrange)):
#print(inrange[i],i)
if tag.ID in inrange[i] and tag.ID not in bin_sets[i]:
px,py=points[i,:]
idx = np.digitize(px, x_bins)-1
idy = np.digitize(py, y_bins)-1
bin_sets[i].add(tag.ID)
bin_lens[idx,idy]+=1
#input()
tag.updatePing(t)
if last_meas+tau<=t:
rate_meas = bin_lens/tau
bin_counts+=1
bin_rates= iterative_average(rate_meas,bin_counts,bin_rates) #iteratively average rate measurement
last_meas=t
bin_sets=[set() for i in range((nx_bins*ny_bins))]
bin_lens=0*bin_lens
t+=time_step
if get_points:
points.shape=(nx_bins,ny_bins,2)
return np.flip(bin_rates,0),np.flip(points,0)
return np.flip(bin_rates,0)
def flow(self, x):
return 0.004*np.array([np.cos(0.25*x[0]), np.sin(0.6*x[1])])
def setMap(self,map=None):
if type(map)==type(None):
self.p=np.array([0.1, 0.1, 0.4, 0.6, 0.2,
0.1, 0.3, 0.5, 0.8, 0.3,
0.2, 0.4, 0.2, 0.2, 0.1,
0.5, 0.9, 0.4, 0.2, 0.1,
0.3, 0.6, 0.3, 0.1, 0.1])
else:
self.p = map
count = 0
self.tag_count = np.zeros_like(self.p)
N=len(self.taglist)
while count < N:
add_prob = np.random.rand()
to_add = add_prob<self.p
for i in range(self.p.size):
if to_add[i] and count<N:
xr,yr,zr = self.abstractPosToRange(i)
mpx,mpy =(self.x_range*1.0/len(self.x_bins),self.y_range*1.0/len(self.y_bins))
self.taglist[count].pos=np.array([xr+mpx*np.random.rand(),yr+mpy*np.random.rand(),zr*np.random.rand()])
self.taglist[count].bin=i
self.tag_count[i]+=1
count+=1
if __name__ == '__main__':
from AcousticTag import AcousticTag
taglist=[]
N=1000
for i in range(N):
taglist.append(AcousticTag(i))
grid = Grid(taglist)
grid.setMap()
tau=.5
# print('tau=',tau)
# print(np.round(grid.groundTruth(tau),decimals=2))
# print('method 2 tau=',tau)
# print(np.round(grid.approximateField(tau,spacing=(200,200),sensorRange=110),decimals=2))
# #print(np.round(grid.groundTruth(tau,fixed=True),decimals=2))
# tau=2
# print('tau=',tau)
# print(np.round(grid.groundTruth(tau),decimals=2))
# print('method 2 tau=',tau)
# print(np.round(grid.approximateField(tau,spacing=(200,200),sensorRange=110),decimals=2))
# #print(np.round(grid.groundTruth(tau,fixed=True),decimals=2))
# tau=5
# print('tau=',tau)
# print(np.round(grid.groundTruth(tau),decimals=2))
# print('method 2 tau=',tau)
# print(np.round(grid.approximateField(tau,spacing=(200,200),sensorRange=100),decimals=2))
# #print(np.round(grid.groundTruth(tau,fixed=True),decimals=2))
# tau=10
# print('tau=',tau)
# print(np.round(grid.groundTruth(tau),decimals=2))
# print('method 2 tau=',tau)
# print(np.round(grid.approximateField(tau,spacing=(200,200),sensorRange=100),decimals=2))
# #print(np.round(grid.groundTruth(tau,fixed=True),decimals=2))
# tau=17
# print('tau=',tau)
# print(np.round(grid.groundTruth(tau,time_step=5),decimals=2))
# print('method 2 tau=',tau)
# print(np.round(grid.approximateField(tau,spacing=(200,200),sensorRange=100),decimals=2))
# #print(np.round(grid.groundTruth(tau,fixed=True,time_step=5),decimals=2))
# tau=34
# print('tau=',tau)
# print(np.round(grid.groundTruth(tau,time_step=5),decimals=2))
# #print(np.round(grid.groundTruth(tau,fixed=True,time_step=5),decimals=2))
# print('method 2 tau=',tau)
# print(np.round(grid.approximateField(tau,spacing=(200,200),sensorRange=100),decimals=2))
tau=2
print(np.round(grid.approximateField(tau,spacing=(200,200),sensorRange=50),decimals=2))