def test_solving_trivial_two_var_problem():
bip = Knapsack(capacity=3)
bip.add_var(Var(), weight=1, value=1)
bip.add_var(Var(), weight=1, value=1)
bip.solve()
assert all([v.value == 1 for v in bip.variables])
assert bip.optimal_value == 2
def test_knapsack_with_negative_values():
bip = Knapsack(capacity=10)
bip.add_var(Var(), weight=4, value=-2)
bip.add_var(Var(), weight=20, value=5)
bip.solve()
assert [v.value for v in bip.variables] == [0, 0]
assert bip.optimal_value == 0
def test_knapsack_with_negative_values_neg_weights():
bip = Knapsack(capacity=10)
x, y = Var(), Var()
bip.add_var(x, weight=15, value=-2)
bip.add_var(y, weight=-5, value=5)
bip.solve()
assert [x.value, y.value] == [0, 1]
assert bip.optimal_value == 5
def test_solving_nontrivial_two_var_problem():
bip = Knapsack(capacity=3)
bip.add_var(Var(), weight=4, value=5)
bip.add_var(Var(), weight=1, value=1)
bip.solve()
assert bip.variables[0].value == 0
assert bip.variables[1].value == 1
assert bip.optimal_value == 1
def test_solving_two_var_problem_with_negative_weights():
bip = Knapsack(capacity=3)
x, y = Var(), Var()
bip.add_var(x, weight=-4, value=5)
bip.add_var(y, weight=1, value=1)
bip.solve()
assert x.value == 1
assert y.value == 1
assert bip.optimal_value == 6
def test_solving_nontrivial_eight_var_problem():
bip = Knapsack(capacity=104)
for weight, value in zip([25, 35, 45, 5, 25, 3, 2, 2],
[350, 400, 450, 20, 70, 8, 5, 5]):
bip.add_var(Var(), weight=weight, value=value)
bip.solve()
values = [v.value for v in bip.variables]
assert values == [1, 0, 1, 1, 1, 0, 1, 1]
assert bip.optimal_value == 900
def test_solving_nontrivial_four_var_problem():
bip = Knapsack(capacity=14)
bip.add_var(Var(), weight=5, value=8)
bip.add_var(Var(), weight=7, value=11)
bip.add_var(Var(), weight=4, value=6)
bip.add_var(Var(), weight=3, value=4)
bip.solve()
values = [v.value for v in bip.variables]
assert values == [0, 1, 1, 1]
assert bip.optimal_value == 21
def test_solving_trivial_problem_alt_init():
bip = Knapsack(capacity=3, weights=[1], values=[1])
bip.solve()
assert bip.variables[0].value == 1
assert bip.optimal_value == 1