def simple_ite_optim(self):
# Create random input and output data
#xL = torch.ones((1, 2), dtype=torch.float16, requires_grad=True)
#xH = torch.ones((1, 2), dtype=torch.float16, requires_grad=True)
xL = torch.tensor([-10.0])
xH = torch.tensor([1.0])
var_map = {'x': 0, 'c': 1}
# x in [-10, 1.0]
# c approaches 0 but c < 0
program = StatementBlock([
# if c <= 0
IfThenElse(
IntervalBool(torch.tensor([0.0, 1.0]), torch.tensor([0.0,
0.0])),
# x = 2x + c
AssignStatement(torch.tensor([[2.0, 1.0], [0.0, 1.0]]),
torch.tensor([0.0, 0.0])),
# x = 700
AssignStatement(torch.tensor([[0.0, 0.0], [0.0, 1.0]]),
torch.tensor([700.0, 0.0]))),
# return x
ReturnStatement(torch.tensor([1.0, 0.0]), torch.tensor([0.0, 0.0]))
])
result_c = optimize(xL, xH, [InferParameter(var_map['c'])], program)[0]
self.assertTrue(result_c >= 0.0)
self.assertTrue(result_c <= 0.25)
def test_ite_bad_ex(self):
val = AbsInterval(torch.tensor([-10.0, 0.0]),
torch.tensor([1.0, -0.0001]))
program = StatementBlock([
IfThenElse(
IntervalBool(torch.tensor([0.0, 1.0]), torch.tensor([0.0,
0.0])),
AssignStatement(torch.tensor([[2.0, 1.0], [0.0, 1.0]]),
torch.tensor([0.0, 0.0])),
AssignStatement(torch.tensor([[0.0, 0.0], [0.0, 1.0]]),
torch.tensor([700.0, 0.0]))),
ReturnStatement(torch.tensor([1.0, 0.0]), torch.tensor([0.0, 0.0]))
])
def test_ite_alwaystrue(self):
# x in [-1, 5]
x = AbsInterval(torch.tensor([-1.0]), torch.tensor([5.0]), 1.0)
# if 0x + 3 >= 0
# then x = 2x
# else x = -5
program = IfThenElse(
IntervalBool(torch.tensor([0.0]), torch.tensor([3.0])),
AssignStatement(torch.tensor([[2.0]]), torch.tensor([0.0])),
AssignStatement(torch.tensor([[0.0]]), torch.tensor([-5.0])))
# x in [-2, 10]
self.assertEqual(
program.propagate(x, t_os),
AbsInterval(torch.tensor([-2.0]), torch.tensor([10.0]), 1.0))
def test_simple_ite(self):
# x in [-1, 5]
x = AbsInterval(torch.tensor([-1.0]), torch.tensor([5.0]), 1.0)
# if x > c
# then x = 2x
# else x = -0.5
program = IfThenElse(
IntervalBool(torch.tensor([1.0]), torch.tensor([0.0])),
AssignStatement(torch.tensor([[2.0]]), torch.tensor([0.0])),
AssignStatement(torch.tensor([[0.0]]), torch.tensor([-0.5])))
# x in [-0.5, 10]
self.assertEqual(
program.propagate(x, t_os),
AbsInterval(torch.tensor([-0.5]), torch.tensor([10.0]), 1.0))
def test_ite_alwaysfalse(self):
# x in [-5, 5]
x = AbsInterval(torch.tensor([-5.0]), torch.tensor([5.0]), 1.0)
# if 2x - 11 >= 0
# then x = 2x
# else x = -5
program = IfThenElse(
IntervalBool(torch.tensor([2.0]), torch.tensor([-11.0])),
AssignStatement(torch.tensor([[2.0]]), torch.tensor([0.0])),
AssignStatement(torch.tensor([[0.0]]), torch.tensor([-5.0])))
# x in [-5, -5]
self.assertEqual(
program.propagate(x, t_os),
AbsInterval(torch.tensor([-5.0]), torch.tensor([-5.0]), 1.0))
# Discontinuity exhibit 1: as input bound increases by 1, output jumps by 17
# x in [-5, 6]
x = AbsInterval(torch.tensor([-5.0]), torch.tensor([6.0]), 1.0)
# x in [-5, 12]
self.assertEqual(
program.propagate(x, t_os),
AbsInterval(torch.tensor([-5.0]), torch.tensor([12.0]), 1.0))
def basic_example():
# Create random input and output data
#xL = torch.ones((1, 2), dtype=torch.float16, requires_grad=True)
#xH = torch.ones((1, 2), dtype=torch.float16, requires_grad=True)
xL = torch.tensor([float(-10.0)])
xH = torch.tensor([float(1.0)])
var_list = ['x', 'c']
diag = torch.diag(torch.ones(len(var_list)))
var_map = {v: k for (k, v) in enumerate(var_list)}
var_b = {v: diag[k] for (k, v) in enumerate(var_list)}
# x in [-10, 1.0]
# c approaches 0 but c < 0
program = StatementBlock([
# if c >= 0
IfThenElse(
IntervalBool(var_b['c'], 0.0),
# x = 2x + c
AssignStatement(
var_map['x'],
ConstExpr(2.0) * VarExpr(var_map['x']) +
VarExpr(var_map['c'])),
# x = 700
AssignStatement(var_map['x'], ConstExpr(700.0))),
# return x
ReturnStatement(VarExpr(var_map['x']))
])
#program = Program([var_map["lin"], var_map["ltarget"]],
# StatementBlock([
# InferAssignStatement(0, )
# ])))
#repl(program)
print("Arguments are found to be: ")
res = optimizeNM(xL, xH, program, 1)
print("Result:")
print(res)
def test_simple_ite_full(self):
x1 = AbsInterval(torch.tensor([-10.0, 0.5]), torch.tensor([1.0, 0.5]),
1.0)
x2 = AbsInterval(torch.tensor([-10.0, -0.5]),
torch.tensor([1.0, -0.5]), 1.0)
program = StatementBlock([
IfThenElse(
IntervalBool(torch.tensor([0.0, 1.0]), torch.tensor([0.0,
0.0])),
AssignStatement(torch.tensor([[2.0, -1.0], [0.0, 1.0]]),
torch.tensor([0.0, 0.0])),
AssignStatement(torch.tensor([[0.0, 0.0], [0.0, 1.0]]),
torch.tensor([700.0, 0.0]))),
ReturnStatement(torch.tensor([1.0, 0.0]), torch.tensor([0.0, 0.0]))
])
self.assertEqual(
program.propagate(x1, t_os),
AbsInterval(torch.tensor([-20.5, 0.5]), torch.tensor([1.5, 0.5]),
1.0))
self.assertEqual(
program.propagate(x2, t_os),
AbsInterval(torch.tensor([700.0, -0.5]),
torch.tensor([700.0, -0.5]), 1.0))
def therm_example():
num_infer = 7
num_input = 2
var_list = ['lin', 'ltarget', "i", "isOn", "K", "curL", 'h', 'tOn', 'tOff']
xL = torch.ones(num_input) * -1.0
xH = torch.ones(num_input) * 10.0
diag = torch.diag(torch.ones(len(var_list)))
var_map = {v: k for (k, v) in enumerate(var_list)}
var_b = {v: diag[k] for (k, v) in enumerate(var_list)}
print(var_map)
program = StatementBlock([
# isOn = 0.0;
AssignStatement(var_map['isOn'], ConstExpr(0.0)),
# K = 0.1;
AssignStatement(var_map['K'], ConstExpr(0.1)),
# curL = lin;
AssignStatement(var_map['curL'], VarExpr(var_map['lin'])),
# assert(tOn < tOff) => assert(tOn - tOff <= 0) => assert(tOff - tOn >= 0)
AssertStatement(IntervalBool(var_b['tOff'] - var_b['tOn'], 0)),
# assert(h >= 0)
AssertStatement(IntervalBool(var_b['h'], 0)),
# assert(h < 20) => assert(-h + 20 >= 0)
AssertStatement(IntervalBool(-var_b['h'], 20.0)),
# if isOn >= 0.5 ==> isOn - 0.5 >= 0
IfThenElse(
IntervalBool(var_b['isOn'], -0.5),
StatementBlock([
# curL = curL + h - K * (curL - lin)
AssignStatement(
var_map['curL'],
VarExpr(var_map['curL']) + VarExpr(var_map['h']) -
VarExpr(var_map['K']) *
(VarExpr(var_map['curL']) - VarExpr(var_map['lin']))),
# if (curL - tOff >= 0.0)
IfThenElse(
IntervalBool(var_b['curL'] - var_b['tOff'], 0.0),
# isOn = 0.0
AssignStatement(var_map['isOn'], ConstExpr(0.0)),
NoOp())
]),
StatementBlock([
# curL = curL - K * (curL - lin)
AssignStatement(
var_map['curL'],
VarExpr(var_map['curL']) - VarExpr(var_map['K']) *
(VarExpr(var_map['curL']) - VarExpr(var_map['lin']))),
# if (-curL + tOn >= 0)
IfThenElse(
IntervalBool(-var_b['curL'] + var_b['tOn'], 0.0),
# isOn = 1.0
AssignStatement(var_map['isOn'], ConstExpr(1.0)),
NoOp())
])),
# assert (-curL + 120 > 0)
AssertStatement(IntervalBool(-var_b['curL'], 120.0)),
# return curL - ltarget
ReturnStatement(
VarExpr(var_map['curL']) - VarExpr(var_map['ltarget']))
])
#infer = torch.tensor([5.0, -5.0, 5.0, 0.0, 0.1, 1.0])
#print("xL: %s, xH: %s, infer: %s" % (str(xL.size()), str(xH.size()), str(infer.size())))
print("Arguments are found to be: ")
print(optimizeNM(xL, xH, program, num_infer))