-
Notifications
You must be signed in to change notification settings - Fork 11
/
utilities.py
289 lines (225 loc) · 7.07 KB
/
utilities.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
import numpy as np
import numpy.random as npr
import scipy.special as sp
import scipy.stats as ss
dpi = 2*np.pi
def bin_vec_to_dec(bin_vec, nsam=1):
"""
This function takes a 1 dim array of zeros and ones and returns the
decimal representation of that array. If nsam > 1, the function operates
on each row, assuming bin_vec is 2 dim matrix with bin_vec.shape[0] = nsam
Parameters
----------
bin_vec : np.array
entries of array are either 0 or 1
nsam : int
number of samples
Returns
-------
int | np.array
"""
def fun(bv):
size = len(bv)
x = np.sum(np.array(
[bv[k]*(1 << (size-1-k)) for k in range(size)]
))
return x
if nsam == 1:
return fun(bin_vec)
elif nsam > 1:
return np.array([fun(bin_vec[sam, :]) for sam in range(nsam)])
else:
assert False, "illegal value for number of samples"
def dec_to_bin_vec(dec, size, nsam=1):
"""
This function takes in an int and it returns a 1 dim array with entries
0 or 1 and length equal to 'size'. It checks to make sure 'size' is
large enough to fit the input. If nsam > 1, the function operates on
each entry, assuming dec is 1 dim matrix with dec.shape[0] = nsam
Parameters
----------
dec : int | np.array
size : int
nsam : int
number of samples
Returns
-------
np.array
entries of array are either 0 or 1
"""
def fun(scalar):
assert scalar < (1 << size), "size " + str(size) + " is too small"\
" to fit bin rep of " + str(scalar)
return np.array([(scalar >> k) & 1 for k in range(size)])
if nsam == 1:
return fun(dec)
elif nsam > 1:
return np.stack([fun(dec[sam]) for sam in range(nsam)])
else:
assert False, "illegal value for number of samples"
def ang_to_cs2_prob(ang, use_sin=1):
"""
This function takes a float 'ang' and returns its cosine or sine
squared, depending on flag 'use_sin'. The function also works if ang is
a numpy array, in which case it acts elementwise and returns an array of
the same shape as 'ang'.
Parameters
----------
ang : float | np.array
use_sin : int
Either 0 or 1
Returns
-------
float | np.array
"""
if use_sin == 1:
fun = np.sin
elif use_sin == 0:
fun = np.cos
else:
assert False
return np.square(fun(ang))
def log_beta_prob(x, conc0, conc1):
"""
This function takes in a probability 'x' and returns the log probability
of 'x', according to the Beta distribution with concentrations 'conc0'
and 'conc1'. The Wikipedia article cited below refers to conc0 as alpha
and to conc1 as beta.
x, conc0, conc1, g0 and g1 are all floats, or they are all numpy arrays
of the same shape. This method works elementwise for arrays.
References
----------
https://en.wikipedia.org/wiki/Beta_distribution
Parameters
----------
x : float | np.array
conc0 : float | np.array
Concentration 0, must be >= 0
conc1 : float | np.array
Concentration 1, must be >= 0
Returns
-------
float | np.array
"""
x = np.clip(x, .00001, .99999)
return np.log(ss.beta.pdf(x, conc0, conc1))
def grad_log_beta_prob(x, conc0, conc1):
"""
This function takes in a probability 'x' and returns the GRADIENT of the
log probability of 'x', according to the Beta distribution with
concentrations 'conc0' and 'conc1'. The Wikipedia article cited below
refers to conc0 as alpha and to conc1 as beta.
x, conc0, conc1, g0, g1 are all floats, or they are all numpy arrays of
the same shape. This method works elementwise for arrays.
References
----------
https://en.wikipedia.org/wiki/Beta_distribution
Parameters
----------
x : float | np.array
x in interval [0, 1]
conc0 : float | np.array
Concentration 0, must be >= 0
conc1 : float | np.array
Concentration 1, must be >= 0
Returns
-------
[g0, g1]: list[float | np.array, float | np.array]
g0 and g1 are tha gradients with respect to the concentrations conc0
and conc1.
"""
def fun(z):
a = np.clip(np.real(z), 1e-5, 15)
b = np.clip(np.imag(z), 1e-5, 15)
return sp.digamma(a+b) - sp.digamma(a)
vfun = np.vectorize(fun)
x = np.clip(x, .00001, .99999)
g0 = np.log(x) + vfun(conc0 + 1j*conc1)
g1 = np.log(1-x) + vfun(conc1 + 1j*conc0)
return [g0, g1]
def log_bern_prob(x, prob1):
"""
This function takes in x= 0 or 1 and returns the log probability of 'x',
according to the Bernoulli distribution. So it returns P(x=0)= 1-prob1,
P(x=1)= prob1
x, prob1 and output are all scalars, or they are all numpy arrays of the
same shape. This method works elementwise for arrays.
References
----------
https://en.wikipedia.org/wiki/Bernoulli_distribution
Parameters
----------
x : int | np.array
either 0 or 1
prob1 : float | np.array
probability in closed interval [0, 1]
Returns
-------
float | np.array
"""
prob1 = np.clip(prob1, .00001, .99999)
return np.log(ss.bernoulli.pmf(x, prob1))
def av_each_elem_in_array_list(list1):
"""
'list1' is a list of numpy arrays of possibly different shapes. This
function takes 'list1' and returns a new list which replaces each entry
of 'list1' by its average (a scalar).
Parameters
----------
list1 : list[np.array]
Returns
-------
list[float]
"""
return np.array([np.mean(x) for x in list1])
def get_shapes_of_array_list(list1):
"""
'list1' is a list of numpy arrays of possibly different shapes. This
function takes 'list1' and returns a new list which replaces each entry
of 'list1' by its shape.
Parameters
----------
list1 : list[np.array]
Returns
-------
list[tuples]
"""
return [arr.shape for arr in list1]
def new_uniform_array_list(val, shapes):
"""
This function returns a list whose kth entry is an array with shape=
shapes[k] and with all entries equal to 'val'.
Parameters
----------
val : float
shapes : list[tuples]
Returns
-------
list[np.array]
"""
return [np.ones(shapes[k])*val for k in range(len(shapes))]
def new_uniformly_random_array_list(low, high, shapes):
"""
This function returns a list whose kth entry is an array with shape=
shapes[k] and with entries selected uniformly at random from
the interval [low, high].
Parameters
----------
low : float
high : float
shapes : list[tuples]
Returns
-------
list[np.array]
"""
return [npr.uniform(low, high, size=shapes[k])
for k in range(len(shapes))]
if __name__ == "__main__":
def main():
vec = dec_to_bin_vec(10, 5)
dec = bin_vec_to_dec(vec)
print("10, vec, dec\n", 10, vec, dec)
dec_to_bin_vec(10, 4)
# uncomment this to see that assert on size works
# dec_to_bin_vec(10, 3)
main()