-
Notifications
You must be signed in to change notification settings - Fork 40
/
t_constraints.py
414 lines (365 loc) · 15.5 KB
/
t_constraints.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
"Unit tests for Constraint, MonomialEquality and SignomialInequality"
import unittest
import numpy as np
from gpkit import Variable, SignomialsEnabled, Posynomial, VectorVariable
from gpkit.nomials import SignomialInequality, PosynomialInequality
from gpkit.nomials import MonomialEquality
from gpkit import Model, ConstraintSet
from gpkit.constraints.costed import CostedConstraintSet
from gpkit.constraints.tight import Tight
from gpkit.constraints.loose import Loose
from gpkit.tests.helpers import run_tests
from gpkit.exceptions import (InvalidGPConstraint, PrimalInfeasible,
DimensionalityError)
from gpkit.constraints.relax import (ConstraintsRelaxed, ConstantsRelaxed,
ConstraintsRelaxedEqually)
from gpkit.constraints.bounded import Bounded
from gpkit.globals import NamedVariables
import gpkit
class TestConstraint(unittest.TestCase):
"Tests for Constraint class"
def test_uninited_element(self):
x = Variable("x")
class SelfPass(Model):
"A model which contains itself!"
def setup(self):
ConstraintSet([self, x <= 1])
self.assertRaises(ValueError, SelfPass)
def test_bad_elements(self):
x = Variable("x")
with self.assertRaises(ValueError):
_ = Model(x, [x == "A"])
with self.assertRaises(ValueError):
_ = Model(x, [x >= 1, x == "A"])
with self.assertRaises(ValueError):
_ = Model(x, [x >= 1, x == "A", x >= 1, ])
with self.assertRaises(ValueError):
_ = Model(x, [x == "A", x >= 1])
v = VectorVariable(2, "v")
with self.assertRaises(ValueError):
_ = Model(x, [v == "A"])
with self.assertRaises(TypeError):
_ = Model(x, [v <= ["A", "B"]])
with self.assertRaises(TypeError):
_ = Model(x, [v >= ["A", "B"]])
def test_evalfn(self):
x = Variable("x")
x2 = Variable("x^2", evalfn=lambda solv: solv[x]**2)
m = Model(x, [x >= 2])
m.unique_varkeys = set([x2.key])
sol = m.solve(verbosity=0)
self.assertAlmostEqual(sol(x2), sol(x)**2)
def test_relax_list(self):
x = Variable("x")
x_max = Variable("x_max", 1)
x_min = Variable("x_min", 2)
constraints = [x_min <= x, x <= x_max]
ConstraintsRelaxed(constraints)
ConstantsRelaxed(constraints)
ConstraintsRelaxedEqually(constraints)
def test_relax_linked(self):
x = Variable("x")
x_max = Variable("x_max", 1)
x_min = Variable("x_min", lambda c: 2*c[x_max])
zero = Variable("zero", lambda c: 0*c[x_max])
constraints = ConstraintSet([x_min + zero <= x, x + zero <= x_max])
_ = ConstantsRelaxed(constraints)
NamedVariables.reset_modelnumbers()
include_min = ConstantsRelaxed(constraints, include_only=["x_min"])
NamedVariables.reset_modelnumbers()
exclude_max = ConstantsRelaxed(constraints, exclude=["x_max"])
self.assertEqual(str(include_min), str(exclude_max))
def test_equality_relaxation(self):
x = Variable("x")
m = Model(x, [x == 3, x == 4])
rc = ConstraintsRelaxed(m)
m2 = Model(rc.relaxvars.prod() * x**0.01, rc)
self.assertAlmostEqual(m2.solve(verbosity=0)(x), 3, places=3)
def test_constraintget(self):
x = Variable("x")
x_ = Variable("x", lineage=[("_", 0)])
xv = VectorVariable(2, "x")
xv_ = VectorVariable(2, "x", lineage=[("_", 0)])
self.assertEqual(Model(x, [x >= 1])["x"], x)
with self.assertRaises(ValueError):
_ = Model(x, [x >= 1, x_ >= 1])["x"]
with self.assertRaises(ValueError):
_ = Model(x, [x >= 1, xv >= 1])["x"]
self.assertTrue(all(Model(xv.prod(), [xv >= 1])["x"] == xv))
with self.assertRaises(ValueError):
_ = Model(xv.prod(), [xv >= 1, xv_ >= 1])["x"]
with self.assertRaises(ValueError):
_ = Model(xv.prod(), [xv >= 1, x_ >= 1])["x"]
def test_additive_scalar(self):
"Make sure additive scalars simplify properly"
x = Variable('x')
c1 = 1 >= 10*x
c2 = 1 >= 5*x + 0.5
self.assertEqual(type(c1), PosynomialInequality)
self.assertEqual(type(c2), PosynomialInequality)
c1hmap, = c1.as_hmapslt1({})
c2hmap, = c2.as_hmapslt1({})
self.assertEqual(c1hmap, c2hmap)
def test_additive_scalar_gt1(self):
"1 can't be greater than (1 + something positive)"
x = Variable('x')
def constr():
"method that should raise a ValueError"
return 1 >= 5*x + 1.1
self.assertRaises(PrimalInfeasible, constr)
def test_init(self):
"Test Constraint __init__"
x = Variable('x')
y = Variable('y')
c = PosynomialInequality(x, ">=", y**2)
self.assertEqual(c.as_hmapslt1({}), [(y**2/x).hmap])
self.assertEqual(c.left, x)
self.assertEqual(c.right, y**2)
self.assertTrue(">=" in str(c))
c = PosynomialInequality(x, "<=", y**2)
self.assertEqual(c.as_hmapslt1({}), [(x/y**2).hmap])
self.assertEqual(c.left, x)
self.assertEqual(c.right, y**2)
self.assertTrue("<=" in str(c))
self.assertEqual(type((1 >= x).latex()), str)
def test_oper_overload(self):
"Test Constraint initialization by operator overloading"
x = Variable('x')
y = Variable('y')
c = (y >= 1 + x**2)
self.assertEqual(c.as_hmapslt1({}), [(1/y + x**2/y).hmap])
self.assertEqual(c.left, y)
self.assertEqual(c.right, 1 + x**2)
self.assertTrue(">=" in str(c))
# same constraint, switched operator direction
c2 = (1 + x**2 <= y) # same as c
self.assertEqual(c2.as_hmapslt1({}), c.as_hmapslt1({}))
def test_sub_tol(self):
" Test PosyIneq feasibility tolerance under substitutions"
x = Variable('x')
y = Variable('y')
z = Variable('z')
PosynomialInequality.feastol = 1e-5
m = Model(z, [x == z, x >= y], {x: 1, y: 1.0001})
self.assertRaises(PrimalInfeasible, m.solve, verbosity=0)
PosynomialInequality.feastol = 1e-3
self.assertEqual(m.substitutions('x'), m.solve(verbosity=0)('x'))
class TestCostedConstraint(unittest.TestCase):
"Tests for Costed Constraint class"
def test_vector_cost(self):
x = VectorVariable(2, "x")
self.assertRaises(ValueError, CostedConstraintSet, x, [])
_ = CostedConstraintSet(np.array(x[0]), [])
class TestMonomialEquality(unittest.TestCase):
"Test monomial equality constraint class"
def test_init(self):
"Test initialization via both operator overloading and __init__"
x = Variable('x')
y = Variable('y')
mono = y**2/x
# operator overloading
mec = (x == y**2)
# __init__
mec2 = MonomialEquality(x, y**2)
self.assertTrue(mono.hmap in mec.as_hmapslt1({}))
self.assertTrue(mono.hmap in mec2.as_hmapslt1({}))
x = Variable("x", "ft")
y = Variable("y")
if gpkit.units:
self.assertRaises(DimensionalityError, MonomialEquality, x, y)
self.assertRaises(DimensionalityError, MonomialEquality, y, x)
def test_vector(self):
"Monomial Equalities with VectorVariables"
x = VectorVariable(3, "x")
self.assertFalse(x == 3)
self.assertTrue(x == x) # pylint: disable=comparison-with-itself
def test_inheritance(self):
"Make sure MonomialEquality inherits from the right things"
F = Variable('F')
m = Variable('m')
a = Variable('a')
mec = (F == m*a)
self.assertTrue(isinstance(mec, MonomialEquality))
def test_non_monomial(self):
"Try to initialize a MonomialEquality with non-monomial args"
x = Variable('x')
y = Variable('y')
def constr():
"method that should raise a TypeError"
MonomialEquality(x*y, x+y)
self.assertRaises(TypeError, constr)
def test_str(self):
"Test that MonomialEquality.__str__ returns a string"
x = Variable('x')
y = Variable('y')
mec = (x == y)
self.assertEqual(type(mec.str_without()), str)
def test_united_dimensionless(self):
"Check dimensionless unit-ed variables work"
x = Variable('x')
y = Variable('y', 'hr/day')
c = MonomialEquality(x, y)
self.assertTrue(isinstance(c, MonomialEquality))
class TestSignomialInequality(unittest.TestCase):
"Test Signomial constraints"
def test_becomes_posy_sensitivities(self):
# pylint: disable=invalid-name
# model from #1165
ujet = Variable("ujet")
PK = Variable("PK")
Dp = Variable("Dp", 0.662)
fBLI = Variable("fBLI", 0.4)
fsurf = Variable("fsurf", 0.836)
mdot = Variable("mdot", 1/0.7376)
with SignomialsEnabled():
m = Model(PK, [mdot*ujet + fBLI*Dp >= 1,
PK >= 0.5*mdot*ujet*(2 + ujet) + fBLI*fsurf*Dp])
var_senss = m.solve(verbosity=0)["sensitivities"]["variables"]
self.assertAlmostEqual(var_senss[Dp], -0.16, 2)
self.assertAlmostEqual(var_senss[fBLI], -0.16, 2)
self.assertAlmostEqual(var_senss[fsurf], 0.19, 2)
self.assertAlmostEqual(var_senss[mdot], -0.17, 2)
# Linked variable
Dp = Variable("Dp", 0.662)
mDp = Variable("-Dp", lambda c: -c[Dp])
fBLI = Variable("fBLI", 0.4)
fsurf = Variable("fsurf", 0.836)
mdot = Variable("mdot", 1/0.7376)
m = Model(PK, [mdot*ujet >= 1 + fBLI*mDp,
PK >= 0.5*mdot*ujet*(2 + ujet) + fBLI*fsurf*Dp])
var_senss = m.solve(verbosity=0)["sensitivities"]["variables"]
self.assertAlmostEqual(var_senss[Dp], -0.16, 2)
self.assertAlmostEqual(var_senss[fBLI], -0.16, 2)
self.assertAlmostEqual(var_senss[fsurf], 0.19, 2)
self.assertAlmostEqual(var_senss[mdot], -0.17, 2)
# fixed negative variable
Dp = Variable("Dp", 0.662)
mDp = Variable("-Dp", -0.662)
fBLI = Variable("fBLI", 0.4)
fsurf = Variable("fsurf", 0.836)
mdot = Variable("mdot", 1/0.7376)
m = Model(PK, [mdot*ujet >= 1 + fBLI*mDp,
PK >= 0.5*mdot*ujet*(2 + ujet) + fBLI*fsurf*Dp])
var_senss = m.solve(verbosity=0)["sensitivities"]["variables"]
self.assertAlmostEqual(var_senss[Dp] + var_senss[mDp], -0.16, 2)
self.assertAlmostEqual(var_senss[fBLI], -0.16, 2)
self.assertAlmostEqual(var_senss[fsurf], 0.19, 2)
self.assertAlmostEqual(var_senss[mdot], -0.17, 2)
def test_init(self):
"Test initialization and types"
D = Variable('D', units="N")
x1, x2, x3 = (Variable("x_%s" % i, units="N") for i in range(3))
with self.assertRaises(TypeError):
sc = (D >= x1 + x2 - x3)
with SignomialsEnabled():
sc = (D >= x1 + x2 - x3)
self.assertTrue(isinstance(sc, SignomialInequality))
self.assertFalse(isinstance(sc, Posynomial))
def test_posyslt1(self):
x = Variable("x")
y = Variable("y")
with SignomialsEnabled():
sc = (x + y >= x*y)
# make sure that the error type doesn't change on our users
with self.assertRaises(InvalidGPConstraint):
_ = sc.as_hmapslt1({})
class TestLoose(unittest.TestCase):
"Test loose constraint set"
def test_raiseerror(self):
x = Variable('x')
x_min = Variable('x_{min}', 2)
m = Model(x, [Loose([x >= x_min]),
x >= 1])
Loose.raiseerror = True
self.assertRaises(RuntimeWarning, m.solve, verbosity=0)
Loose.raiseerror = False
def test_posyconstr_in_gp(self):
"Tests loose constraint set with solve()"
x = Variable('x')
x_min = Variable('x_{min}', 2)
m = Model(x, [Loose([x >= x_min]),
x >= 1])
sol = m.solve(verbosity=0)
self.assertIs(
sol["warnings"]["Unexpectedly Tight Constraints"][0][1], m[0][0])
self.assertAlmostEqual(m[0][0].relax_sensitivity, +1, 3)
m.substitutions[x_min] = 0.5
self.assertAlmostEqual(m.solve(verbosity=0)["cost"], 1)
def test_posyconstr_in_sp(self):
x = Variable('x')
y = Variable('y')
x_min = Variable('x_min', 1)
y_min = Variable('y_min', 2)
with SignomialsEnabled():
sig_constraint = (x + y >= 3.5)
m = Model(x*y, [Loose([x >= y]),
x >= x_min, y >= y_min, sig_constraint])
sol = m.localsolve(verbosity=0)
self.assertIs(
sol["warnings"]["Unexpectedly Tight Constraints"][0][1], m[0][0])
self.assertAlmostEqual(m[0][0].relax_sensitivity, +1, 3)
m.substitutions[x_min] = 2
m.substitutions[y_min] = 1
self.assertAlmostEqual(m.localsolve(verbosity=0)["cost"], 2.5, 5)
class TestTight(unittest.TestCase):
"Test tight constraint set"
def test_posyconstr_in_gp(self):
"Tests tight constraint set with solve()"
x = Variable('x')
x_min = Variable('x_{min}', 2)
m = Model(x, [Tight([x >= 1]),
x >= x_min])
sol = m.solve(verbosity=0)
self.assertIs(
sol["warnings"]["Unexpectedly Loose Constraints"][0][1], m[0][0])
self.assertAlmostEqual(m[0][0].rel_diff, 1, 3)
m.substitutions[x_min] = 0.5
self.assertAlmostEqual(m.solve(verbosity=0)["cost"], 1)
def test_posyconstr_in_sp(self):
x = Variable('x')
y = Variable('y')
with SignomialsEnabled():
sig_constraint = (x + y >= 0.1)
m = Model(x*y, [Tight([x >= y]),
x >= 2, y >= 1, sig_constraint])
sol = m.localsolve(verbosity=0)
self.assertIs(
sol["warnings"]["Unexpectedly Loose Constraints"][0][1], m[0][0])
self.assertAlmostEqual(m[0][0].rel_diff, 1, 3)
m.pop(1)
self.assertAlmostEqual(m.localsolve(verbosity=0)["cost"], 1, 5)
def test_sigconstr_in_sp(self):
"Tests tight constraint set with localsolve()"
x = Variable('x')
y = Variable('y')
x_min = Variable('x_{min}', 2)
y_max = Variable('y_{max}', 0.5)
with SignomialsEnabled():
m = Model(x, [Tight([x + y >= 1]),
x >= x_min,
y <= y_max])
sol = m.localsolve(verbosity=0)
self.assertIs(
sol["warnings"]["Unexpectedly Loose Constraints"][0][1], m[0][0])
self.assertGreater(m[0][0].rel_diff, 0.5)
m.substitutions[x_min] = 0.5
self.assertAlmostEqual(m.localsolve(verbosity=0)["cost"], 0.5, 5)
class TestBounded(unittest.TestCase):
"Test bounded constraint set"
def test_substitution_issue905(self):
x = Variable("x")
y = Variable("y")
m = Model(x, [x >= y], {"y": 1})
bm = Model(m.cost, Bounded(m))
sol = bm.solve(verbosity=0)
self.assertAlmostEqual(sol["cost"], 1.0)
bm = Model(m.cost, Bounded(m, lower=1e-10))
sol = bm.solve(verbosity=0)
self.assertAlmostEqual(sol["cost"], 1.0)
bm = Model(m.cost, Bounded(m, upper=1e10))
sol = bm.solve(verbosity=0)
self.assertAlmostEqual(sol["cost"], 1.0)
TESTS = [TestConstraint, TestMonomialEquality, TestSignomialInequality,
TestTight, TestLoose, TestBounded, TestCostedConstraint]
if __name__ == "__main__": # pragma: no cover
run_tests(TESTS)