-
Notifications
You must be signed in to change notification settings - Fork 0
/
AutomatedTestsPhaseOne.py
422 lines (341 loc) · 11.1 KB
/
AutomatedTestsPhaseOne.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
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
from PhaseOne import *
import numpy as np
try:
import cupy as cp
gpu_flag = True
except Exception:
gpu_flag = False
# Alias
cp = np
print("Not able to run tests with GPU")
def test_phase_one_gradient(use_cupy=False, tol=1e-8):
"""
Tests the function 'phase_one_gradient' with some numerical examples.
-- input: tol - tolerance for if test was passed or not
-- output: int - 1 if test was passed, 0 otherwise
"""
if use_cupy:
print("Testing phase one gradient with cupy")
else:
print("Testing phase one gradient")
# Constants
G = np.array([[1, 2, 3], [4, 5, 6]])
h = np.array([[2, 3]])
x = np.ones(3)
s = np.max(G @ x - h) + 1
t = 1
mu = 15
true_grad_x = np.array([1, 2, 3]) / 9 + np.array([4, 5, 6])
true_grad_s = np.array([-1 / 9])
true_grad = np.hstack([true_grad_x, true_grad_s])
if use_cupy:
true_grad = cp.array(true_grad)
phaseOne = PhaseOneSolver(G, h, mu, use_cupy=use_cupy)
grad = phaseOne.phase_one_gradient(t)
if use_cupy:
if cp.linalg.norm(true_grad - grad) <= tol:
# passed
pass
else:
print("The function 'phase_one_gradient' did not pass the test")
return 0
else:
if np.linalg.norm(true_grad - grad) <= tol:
# passed
pass
else:
print("The function 'phase_one_gradient' did not pass the test")
return 0
G = np.array([[-1, -3], [-1, 1], [1, -2], [1, 4]])
h = np.array([-6, 2, -2, 12])
x = np.array([1, 1])
s = 3
t = 1
mu = 15
true_grad_x = (
np.array([-1, -3])
+ np.array([-1, 1]) / 5
+ np.array([1, -2]) / 2
+ np.array([1, 4]) / 10
)
true_grad_s = -1 / 5 - 1 / 2 - 1 / 10
true_grad = np.hstack([true_grad_x, true_grad_s])
if use_cupy:
true_grad = cp.array(true_grad)
phaseOne = PhaseOneSolver(G, h, mu, use_cupy=use_cupy)
grad = phaseOne.phase_one_gradient(t)
if use_cupy:
if cp.linalg.norm(true_grad - grad) <= tol:
# passed
return 1
else:
print("The function 'phase_one_gradient' did not pass the test")
return 0
else:
if np.linalg.norm(true_grad - grad) <= tol:
# passed
return 1
else:
print("The function 'phase_one_gradient' did not pass the test")
return 0
def test_phase_one_hessian(use_cupy=False, tol=1e-8):
"""
Tests the function 'phase_one_hessian' with some numerical examples.
-- input: tol - tolerance for if test was passed or not
-- output: int - 1 if test was passed, 0 otherwise
"""
if use_cupy:
print("Testing phase one hessian with cupy")
else:
print("Testing phase one hessian")
G = np.array([[1, 2, 3], [4, 5, 6]])
h = np.array([[2, 3]])
x = np.ones(3)
s = np.max(G @ x - h) + 1
mu = 15
true_hess_xx = np.array([[1, 2, 3], [2, 4, 6], [3, 6, 9]]) / 81 + np.array(
[[16, 20, 24], [20, 25, 30], [24, 30, 36]]
)
true_hess_xs = np.reshape(
-np.array([1, 2, 3]) / 81 - np.array([4, 5, 6]), newshape=(3, 1)
)
true_hess_ss = np.array([1 + 1 / 81])
true_hess = np.block([[true_hess_xx, true_hess_xs], [true_hess_xs.T, true_hess_ss]])
if use_cupy:
true_hess = cp.array(true_hess)
phaseOne = PhaseOneSolver(G, h, mu, use_cupy=use_cupy)
hess = phaseOne.phase_one_hessian()
if use_cupy:
if cp.linalg.norm(true_hess - hess) <= tol:
# passed
pass
else:
print("The function 'phase_one_hessian' did not pass the test")
return 0
else:
if np.linalg.norm(true_hess - hess) <= tol:
# passed
pass
else:
print("The function 'phase_one_hessian' did not pass the test")
return 0
G = np.array([[-1, -3], [-1, 1], [1, -2], [1, 4]])
h = np.array([-6, 2, -2, 12])
x = np.array([1, 1])
s = 3
mu = 15
true_hess_xx = (
np.array([[1, 3], [3, 9]])
+ np.array([[1, -1], [-1, 1]]) / 25
+ np.array([[1, -2], [-2, 4]]) / 4
+ np.array([[1, 4], [4, 16]]) / 100
)
true_hess_xs = (
-np.array([-1, -3])
- np.array([-1, 1]) / 25
- np.array([1, -2]) / 4
- np.array([1, 4]) / 100
)
true_hess_xs = np.reshape(true_hess_xs, (-1, 1))
true_hess_ss = 1 + 1 / 25 + 1 / 4 + 1 / 100
true_hess = np.block([[true_hess_xx, true_hess_xs], [true_hess_xs.T, true_hess_ss]])
if use_cupy:
true_hess = cp.array(true_hess)
phaseOne = PhaseOneSolver(G, h, mu, use_cupy=use_cupy)
hess = phaseOne.phase_one_hessian()
if use_cupy:
if cp.linalg.norm(true_hess - hess) <= tol:
# passed again
return 1
else:
print("The function 'phase_one_hessian' did not pass the test")
return 0
else:
if np.linalg.norm(true_hess - hess) <= tol:
# passed again
return 1
else:
print("The function 'phase_one_hessian' did not pass the test")
return 0
def test_phase_one_objective(use_cupy=False, tol=1e-8):
"""
Tests the function 'phase_one_objective_interior' with some numerical examples.
-- input: tol - tolerance for if test was passed or not
-- output: int - 1 if test was passed, 0 otherwise
"""
if use_cupy:
print("Testing phase one objective with cupy")
else:
print("Testing phase one objective")
G = np.array([[1, 2, 3], [4, 5, 6]])
h = np.array([[2, 3]])
if use_cupy:
x = cp.ones(3)
else:
x = np.ones(3)
s = 13
t = 1
mu = 15
true_val = 13 - np.log(9)
phaseOne = PhaseOneSolver(G, h, mu, use_cupy=use_cupy)
val = phaseOne.phase_one_objective(x, s, t)
# Scalar, don't need to change norm
if np.linalg.norm(true_val - val) <= tol:
return 1
else:
print(
"The function 'phase_one_phase_one_objective_interior' did not pass the test"
)
return 0
def test_phase_one(use_cupy=False, linear_solver="solve"):
"""
Tests the function 'phase_one' with some numerical examples.
-- inputs: use_cupy - boolean whether or not to use cupy
linear_solver - which solver to use, 'solve' or 'cg'
-- output: int - 1 if test was passed, 0 otherwise
"""
if use_cupy:
print("Testing phase one with cupy and solver '" + linear_solver + "'")
else:
print("Testing phase one without cupy but with solver '" + linear_solver + "'")
print("Phase one is initialized inside the set")
# phase_one is initialized inside this set
G = np.array([[1, 3], [1, 1], [-1, 0], [0, -1]])
h = np.array([9, 5, 0, 0])
mu = 15
phaseOne = PhaseOneSolver(G, h, mu, use_cupy=use_cupy, linear_solver=linear_solver)
x, s, _ = phaseOne.solve()
if use_cupy:
x = cp.asnumpy(x)
if phaseOne.s < 0 and np.max(G @ x - h) <= 0:
# is ok
pass
else:
print("The function 'phase_one' did not pass the test")
return 0
print("Phase one is initialized outside the set")
# Set which phase_one is not initialized in
G = np.array([[-1, -3], [-1, 1], [-1, 2], [1, 4]])
h = np.array([-6, 2, 2, 12])
mu = 15
phaseOne = PhaseOneSolver(G, h, mu, use_cupy=use_cupy, linear_solver=linear_solver)
x, s, _ = phaseOne.solve()
if use_cupy:
x = cp.asnumpy(x)
if s < 0 and np.max(G @ x - h) <= 0:
# is ok
pass
else:
print("The function 'phase_one' did not pass the test")
return 0
print("Phase one is initialized outside of an unbounded set")
# Unbounded set which phase_one is not initialied in
G = np.array([[1, -2], [-3, 1]])
h = np.array([-2, 0])
mu = 15
phaseOne = PhaseOneSolver(G, h, mu, use_cupy=use_cupy, linear_solver=linear_solver)
x, s, _ = phaseOne.solve()
if use_cupy:
x = cp.asnumpy(x)
if s < 0 and np.max(G @ x - h) <= 0:
# is ok
pass
else:
print("The function 'phase_one' did not pass the test")
return 0
print("Phase one tries to find an empty set")
# Empty set
G = np.array([[3, -1], [-1, 5], [-1, 0], [0, -1]])
h = np.array([-2, 1.5, 0, 0])
mu = 15
phaseOne = PhaseOneSolver(G, h, mu, use_cupy=use_cupy, linear_solver=linear_solver)
x, s, _ = phaseOne.solve()
if use_cupy:
x = cp.asnumpy(x)
if s > 0:
# is ok
pass
else:
print("The function 'phase_one' did not pass the test")
return 0
print("Phase one is given a problem with high dimension")
# Try a problem with larger size
np.random.seed(0)
m, n = 200, 1000
G = np.random.uniform(low=-10, high=10, size=(m, n))
x = np.random.uniform(low=-5, high=5, size=(n))
h = G @ x + 1
mu = 15
phaseOne = PhaseOneSolver(G, h, mu, use_cupy=use_cupy, linear_solver=linear_solver)
x, s, _ = phaseOne.solve()
if use_cupy:
x = cp.asnumpy(x)
if s < 0 and np.max(G @ x - h) < 0:
# is ok
pass
else:
print("The function 'phase_one' did not pass the test")
return 0
return 1
def test_initialized_phase_one(use_cupy=False, linear_solver="solve"):
"""
Tests the function 'phase_one' with an initialized point.
-- inputs: use_cupy - boolean whether or not to use cupy
linear_solver - which solver to use, 'solve' or 'cg'
-- output: int - 1 if test was passed, 0 otherwise
"""
if use_cupy:
print(
"Testing phase one given an intialization with cupy and solver '"
+ linear_solver
+ "'"
)
else:
print(
"Testing phase one given an initialization without cupy with solver '"
+ linear_solver
+ "'"
)
# Set which phase_one is not initialized in
G = np.array([[-1, -3], [-1, 1], [-1, 2], [1, 4]])
h = np.array([-6, 2, 2, 12])
x0 = np.array([-2, -3])
mu = 15
phaseOne = PhaseOneSolver(
G, h, mu, x0=x0, use_cupy=use_cupy, linear_solver=linear_solver
)
x, s, _ = phaseOne.solve()
if use_cupy:
x = cp.asnumpy(x)
if s < 0 and np.max(G @ x - h) <= 0:
# is ok
pass
else:
print("The function 'phase_one' did not pass the test")
return 0
return 1
def automated_tests(use_cupy=True):
tests = [
test_phase_one_objective,
test_phase_one_gradient,
test_phase_one_hessian,
test_initialized_phase_one,
test_phase_one,
]
passed = 0
max_score = 0
print("Start tests")
for count, test in enumerate(tests):
passed += test()
max_score += 1
if use_cupy:
passed += test(use_cupy=True)
max_score += 1
if count >= len(tests) - 2:
passed += test(linear_solver="cg")
max_score += 1
if use_cupy:
passed += test(use_cupy=True, linear_solver="cg")
max_score += 1
print("Automated tests are done.")
print("Passed {} / {} tests".format(passed, max_score))