def test_same_values(self):
print("_____Test Same Values_____")
# constant value, and the purities form a uniform random distribution
value = 10
max_weight = 100
for n_elems in range(self.min_n_elems, self.max_n_elems + 1):
# values = [np.random.exponential() for _ in range(n_elems)]
weights = [random.randrange(start=1, stop=max_weight) for _ in range(n_elems)]
# print("values={values}".format(values=values))
values = [value] * n_elems
# print("weights={weights}".format(weights=weights))
W = np.sum(weights) / 3
optimum_value = sum(bruteforce.knapsack(values, weights, W)[0])
greedy_value = sum(greedy.knapsack(values, weights, W)[0])
if greedy_value == 0 and optimum_value == 0:
continue
optimality = greedy_value / optimum_value
print(
"n_elems={n_elems} \t\t (gred/opt)=({gred}/{opt}) \t\t optimality={optimality}".format(
n_elems=n_elems, optimality=optimality, gred=greedy_value, opt=optimum_value
)
)
def test_same_weight(self):
print("_____Test Same Weight_____")
# constant weight, and the values form a uniform random distribution
weight = 31
max_value = 100
for n_elems in range(self.min_n_elems, self.max_n_elems + 1):
values = [random.randrange(start=1, stop=max_value) for _ in range(n_elems)]
# print("values={values}".format(values=values))
weights = [weight] * n_elems
# print("weights={weights}".format(weights=weights))
W = np.sum(weights) / 2
optimum_value = sum(bruteforce.knapsack(values, weights, W)[0])
greedy_value = sum(greedy.knapsack(values, weights, W)[0])
if greedy_value == 0 and optimum_value == 0:
continue
optimality = greedy_value / optimum_value
print(
"n_elems={n_elems} \t\t optimality={optimality}\t\t ({gred}/{opt})".format(
n_elems=n_elems, optimality=optimality, gred=greedy_value, opt=optimum_value
)
)
def test_same_weight(self):
print("_____Test Same Weight_____")
# constant weight, and the values form a uniform random distribution
weight = 31
max_value = 100
for n_elems in range(self.min_n_elems, self.max_n_elems + 1):
values = [
random.randrange(start=1, stop=max_value)
for _ in range(n_elems)
]
# print("values={values}".format(values=values))
weights = [weight] * n_elems
# print("weights={weights}".format(weights=weights))
W = np.sum(weights) / 2
optimum_value = sum(bruteforce.knapsack(values, weights, W)[0])
greedy_value = sum(greedy.knapsack(values, weights, W)[0])
if greedy_value == 0 and optimum_value == 0:
continue
optimality = greedy_value / optimum_value
print(
"n_elems={n_elems} \t\t optimality={optimality}\t\t ({gred}/{opt})"
.format(n_elems=n_elems,
optimality=optimality,
gred=greedy_value,
opt=optimum_value))
def test_same_values(self):
print("_____Test Same Values_____")
# constant value, and the purities form a uniform random distribution
value = 10
max_weight = 100
for n_elems in range(self.min_n_elems, self.max_n_elems + 1):
# values = [np.random.exponential() for _ in range(n_elems)]
weights = [
random.randrange(start=1, stop=max_weight)
for _ in range(n_elems)
]
# print("values={values}".format(values=values))
values = [value] * n_elems
# print("weights={weights}".format(weights=weights))
W = np.sum(weights) / 3
optimum_value = sum(bruteforce.knapsack(values, weights, W)[0])
greedy_value = sum(greedy.knapsack(values, weights, W)[0])
if greedy_value == 0 and optimum_value == 0:
continue
optimality = greedy_value / optimum_value
print(
"n_elems={n_elems} \t\t (gred/opt)=({gred}/{opt}) \t\t optimality={optimality}"
.format(n_elems=n_elems,
optimality=optimality,
gred=greedy_value,
opt=optimum_value))
def test_manual_4_same_value(self):
print("_______Test 4: same value _______")
values = [10, 10, 10, 10, 10, 10, 10]
weights = [2, 8, 4, 10, 15, 3, 6]
for W in range(2, sum(weights) + 1):
optimum_value = sum(bruteforce.knapsack(values, weights, W)[0])
greedy_value = sum(greedy.knapsack(values, weights, W)[0])
optimality = greedy_value / optimum_value
print(
"W={W} \t (gred/opt)=({gred}/{opt}) \t optimality={optimality}".format(
W=W, optimality=optimality, gred=greedy_value, opt=optimum_value
)
)
def test_manual_3_same_weight(self):
print("_______Test 3: same weight_______")
values = [29, 28, 21, 17, 16, 13, 12, 10]
weights = [2] * len(values)
for W in range(2, sum(weights) + 1):
optimum_value = sum(bruteforce.knapsack(values, weights, W)[0])
greedy_value = sum(greedy.knapsack(values, weights, W)[0])
optimality = greedy_value / optimum_value
print(
"W={W} \t (gred/opt)=({gred}/{opt}) \t optimality={optimality}".format(
W=W, optimality=optimality, gred=greedy_value, opt=optimum_value
)
)
def test_manual_2(self):
print("_______Test 2_______")
values = [140, 120, 100, 80, 60, 40, 20]
weights = [10, 6, 2, 5, 3, 4, 1]
for W in range(1, sum(weights) + 1):
optimum_value = sum(bruteforce.knapsack(values, weights, W)[0])
greedy_value = sum(greedy.knapsack(values, weights, W)[0])
optimality = greedy_value / optimum_value
print(
"W={W} \t (gred/opt)=({gred}/{opt}) \t optimality={optimality}".format(
W=W, optimality=optimality, gred=greedy_value, opt=optimum_value
)
)
def test_manual_4_same_value(self):
print("_______Test 4: same value _______")
values = [10, 10, 10, 10, 10, 10, 10]
weights = [2, 8, 4, 10, 15, 3, 6]
for W in range(2, sum(weights) + 1):
optimum_value = sum(bruteforce.knapsack(values, weights, W)[0])
greedy_value = sum(greedy.knapsack(values, weights, W)[0])
optimality = greedy_value / optimum_value
print(
"W={W} \t (gred/opt)=({gred}/{opt}) \t optimality={optimality}"
.format(W=W,
optimality=optimality,
gred=greedy_value,
opt=optimum_value))
def test_manual_3_same_weight(self):
print("_______Test 3: same weight_______")
values = [29, 28, 21, 17, 16, 13, 12, 10]
weights = [2] * len(values)
for W in range(2, sum(weights) + 1):
optimum_value = sum(bruteforce.knapsack(values, weights, W)[0])
greedy_value = sum(greedy.knapsack(values, weights, W)[0])
optimality = greedy_value / optimum_value
print(
"W={W} \t (gred/opt)=({gred}/{opt}) \t optimality={optimality}"
.format(W=W,
optimality=optimality,
gred=greedy_value,
opt=optimum_value))
def test_manual_2(self):
print("_______Test 2_______")
values = [140, 120, 100, 80, 60, 40, 20]
weights = [10, 6, 2, 5, 3, 4, 1]
for W in range(1, sum(weights) + 1):
optimum_value = sum(bruteforce.knapsack(values, weights, W)[0])
greedy_value = sum(greedy.knapsack(values, weights, W)[0])
optimality = greedy_value / optimum_value
print(
"W={W} \t (gred/opt)=({gred}/{opt}) \t optimality={optimality}"
.format(W=W,
optimality=optimality,
gred=greedy_value,
opt=optimum_value))