def test_knapsack_unbounded_for_approx(self):
# http://math.stackexchange.com/questions/720001/
# Item (Value, Weight)
i1 = Item (90, 9)
i2 = Item (19, 2)
i3 = Item (1, 1)
self.assertEqual((95,[0,5,0]), knapsack_unbounded([i1,i2,i3], 10))
# approximation chooses this one
self.assertEqual((91,[1,0,1]), knapsack_approximate([i1,i2,i3], 10))
def test_knapsack_unbounded_for_approx(self):
# http://math.stackexchange.com/questions/720001/
# Item (Value, Weight)
i1 = Item (90, 9)
i2 = Item (19, 2)
i3 = Item (1, 1)
self.assertEqual((95,[0,5,0]), knapsack_unbounded([i1,i2,i3], 10))
# approximation chooses this one
self.assertEqual((91,[1,0,1]), knapsack_approximate([i1,i2,i3], 10))
def reportOne():
"""
To see the generates matrices, modify the knapsack methods temporarily to
print out the matrices.
"""
items = [Item(4, 4), Item(8, 8), Item(9, 9), Item(10, 10)]
W = 33
print("Knapsack Unbounded")
w_un = knapsack_unbounded(items[:], W)
print(w_un)
print("Knapsack Approximate")
w_ap = knapsack_approximate(items[:], W)
print(w_ap)
print("Knapsack 0/1")
w_01 = knapsack_01(items[:], W)
print(w_01)
def reportOne():
"""
To see the generates matrices, modify the knapsack methods temporarily to
print out the matrices.
"""
items = [Item(4,4), Item(8,8), Item(9,9), Item(10,10)]
W = 33
print ("Knapsack Unbounded")
w_un = knapsack_unbounded(items[:], W);
print (w_un)
print ("Knapsack Approximate")
w_ap = knapsack_approximate(items[:], W);
print (w_ap)
print ("Knapsack 0/1")
w_01 = knapsack_01(items[:], W);
print (w_01)
def trials():
"""
Locate knapsack items that has different results for :
1) knapsack_unbounded
2) knapsack_01
3) knapsack_approximate
Found this nice example:
W = 33
(33, [6, 0, 0, 1]) (32, [8, 0, 0, 0]) (31, [1, 1, 1, 1])
(val= 4 , weight= 4
(val= 8 , weight= 8
(val= 10 , weight= 10
(val= 9 , weight= 9
"""
# 103 207 4 2 523
for t in range(10000):
items = []
# Item (Value, Weight)
for _ in range(5):
i = random.randint(3, 11)
items.append(Item(i, i))
W = 33
w_un = knapsack_unbounded(items[:], W)
w_ap = knapsack_approximate(items[:], W)
w_01 = knapsack_01(items[:], W)
if w_un[0] != w_ap[0] and w_ap[0] != w_01[0] and w_un[0] != w_01[0]:
print(w_un, w_ap, w_01)
for i in items:
print('(val=', i.value, ', weight=', i.weight)
return
def trials():
"""
Locate knapsack items that has different results for :
1) knapsack_unbounded
2) knapsack_01
3) knapsack_approximate
Found this nice example:
W = 33
(33, [6, 0, 0, 1]) (32, [8, 0, 0, 0]) (31, [1, 1, 1, 1])
(val= 4 , weight= 4
(val= 8 , weight= 8
(val= 10 , weight= 10
(val= 9 , weight= 9
"""
# 103 207 4 2 523
for t in range(10000):
items = []
# Item (Value, Weight)
for _ in range(5):
i = random.randint(3,11)
items.append(Item(i,i))
W = 33
w_un = knapsack_unbounded(items[:], W);
w_ap = knapsack_approximate(items[:], W);
w_01 = knapsack_01(items[:], W);
if w_un[0] != w_ap[0] and w_ap[0] != w_01[0] and w_un[0] != w_01[0]:
print (w_un, w_ap, w_01)
for i in items:
print ('(val=', i.value, ', weight=', i.weight);
return