def test_add_multiple_nodes_left(self):
new_tree = bst(20)
smaller_node = bst(19)
even_smaller_node = bst(18)
new_tree.add(smaller_node)
new_tree.add(even_smaller_node)
self.assertTrue(new_tree.get_left_child().get_left_child().value == even_smaller_node.value)
def __call__(self, x, train):
h = bst.bst(self.b0(self.conv0(x)))
h = bst.bst(self.b1(self.conv1(h)))
h = bst.bst(self.b2(self.conv2(h)))
h = F.average_pooling_2d(h, 48)
h = self.b3(self.fc0(h))
return h
def test_add_multiple_nodes_right(self):
new_tree = bst(20)
larger_node = bst(21)
even_larger_node = bst(22)
new_tree.add(larger_node)
new_tree.add(even_larger_node)
self.assertTrue(new_tree.get_right_child().get_right_child().value == even_larger_node.value)
def huffman(freq_table):
# should use min-heap for pool
pool = [bst.bst((freq_table[key], key)) for key in freq_table.keys()]
while len(pool) > 1:
pool.sort(key=lambda x: -x.data[0])
a, b = pool.pop(), pool.pop()
weight = a.data[0] + b.data[0]
c = bst.bst((weight, ''))
c.left, c.right = a, b
pool.append(c)
return pool[0]
def __call__(self, x, train, batch_size):
h = self.b_conv0(da.dst(self.conv0(x)))
h = self.block0(h, train)
h = bst.bst(self.b_conv1(self.conv1(h)))
h = self.block1(h, train)
h = bst.bst(self.b_conv2(self.conv2(h)))
h = self.block2(h, train)
h = bst.bst(self.b_conv3(self.conv3(h)))
h = self.block3(h, train)
h = bst.bst(self.b_conv4(self.conv4(h)))
h = F.max_pooling_2d(h, 2, 2)
h = F.average_pooling_2d(h, 2, 2, 1)
h = Variable(cupy.reshape(h.data, [batch_size, -1]),
volatile=not train)
h = self.b_dense0(self.fc0(h))
return h
def setUp(self):
self.tree = bst()
self.tree.insert(9)
self.tree.insert(6)
self.tree.insert(12)
self.tree.insert(3)
self.tree.insert(8)
self.tree.insert(10)
self.tree.insert(15)
self.tree.insert(7)
self.tree.insert(18)
def setUp(self):
self.tree = bst()
self.tree.insert(9)
self.tree.insert(6)
self.tree.insert(12)
self.tree.insert(3)
self.tree.insert(8)
self.tree.insert(10)
self.tree.insert(15)
self.tree.insert(7)
self.tree.insert(18)
def _reconcile(self):
if self.NEED_RECONCILE:
self.suspend_watch()
temp = get_file_list(self.DIR_PATH, self.SETTINGS.get_wallpaper_recursive())
temp_tree = bst.bst(temp)
for my_file in self.LOCAL_FILE_LIST:
if temp_tree.extract(my_file) != my_file:
self.list_remove(my_file)
temp_list = temp_tree.as_list()
if len(temp_list) > 0:
self.LOCAL_FILE_LIST += temp_tree.as_list()
self.NEED_SAVE = True
self.instate_watch()
self.NEED_RECONCILE = False
def main():
bst_tree = bst()
# vals = [round(random.random()*100) for i in xrange(15)]
vals = [myString('(')]
bst_tree.add_val( vals[0] )
for i in xrange(5):
nvals = []
while vals:
v = vals.pop()
nvals.append( myString(v + '(') )
bst_tree.add_val( nvals[-1] )
nvals.append( myString(v + ')') )
bst_tree.add_val( nvals[-1] )
vals = nvals
bst_tree.print_bfs()
print "BALANCED LEAVES:", bst_tree.get_blncd_leaves()
def test_check_if_value_doesnt_exist(self):
new_tree = bst(20)
node1 = bst(10)
node2 = bst(25)
node3 = bst(30)
node4 = bst(4)
node5 = bst(89)
new_tree.add(node1)
new_tree.add(node2)
new_tree.add(node3)
new_tree.add(node4)
new_tree.add(node5)
self.assertFalse(new_tree.find(200))
def test():
# 1
print '#'
print 'problme 1: please refer to the tree.pdf file to see the tree as a graph'
print '#'
# node list for construct binary search tree
nd_list = [
65, 28, 22, 46, 35, 32, 40, 48, 47, 55, 83, 78, 89, 85, 86, 91, 93
]
print 'tree node list'
print nd_list
bt = bst()
for i in range(len(nd_list)):
bt.bst_insert(bt.root, nd_list[i])
print 'original tree, in order traversal'
bt.bst_in_trav(bt.root)
bt1 = copy.deepcopy(bt)
bt2 = copy.deepcopy(bt)
print
print 'delete a leaf: (86), in order traversal'
bt.delete_value(bt.root, 86)
bt.bst_in_trav(bt.root)
print
print 'delete (91), its successor is (93), (93) is a leaf, in order traversal'
bt1.delete_value(bt1.root, 91)
bt1.bst_in_trav(bt1.root)
print
print 'delete (83), its successor is (85), (85) is not a leaf, in order traversal'
bt2.delete_value(bt2.root, 83)
bt2.bst_in_trav(bt2.root)
print
print '#'
# 2
# bubble sort
print 'problem 2: bubble sort'
print '#'
print 'original list: '
t_list = [32, 7, 45, 13, 64, 9, 3, 11, 9, 18, 69, 23, 33, 99, 76, 86]
print t_list
print 'bubble sort passes'
bbsort(t_list)
def __call__(self, x):
h1 = bst.bst(self.b1(self.c1(x), test=not self.train))
h2 = bst.bst(self.b2(self.c2(h1), test=not self.train))
h3 = bst.bst(self.b3(self.c3(h2), test=not self.train))
h4 = bst.bst(self.b4(self.l1(h3), test=not self.train))
return self.b5(self.l2(h4), test=not self.train)
def __call__(self, x):
h1 = bst.bst(self.b1(self.l1(x), test=not self.train))
h2 = bst.bst(self.b2(self.l2(h1), test=not self.train))
return self.b3(self.l3(h2), test=not self.train)
def test_for_parent_when_empty(self):
new_tree = bst()
self.assertTrue(new_tree.parent == None)
def test_if_root(self):
new_tree = bst()
self.assertTrue(new_tree.is_root())
def test_parent_of_left(self):
new_tree = bst(20)
smaller_node = bst(19)
new_tree.add(smaller_node)
self.assertTrue(new_tree.get_left_child().get_parent().value == new_tree.value)
def __call__(self, x):
x = x * 256
h1 = bst(self.b1(self.l1(x), test=not self.train))
h2 = bst(self.b2(self.l2(h1), test=not self.train))
return self.b3(self.l3(h2), test=not self.train)
def test_for_left_when_empty(self):
new_tree = bst()
self.assertFalse(new_tree.has_left_child())
def test_for_both_when_empty(self):
new_tree = bst()
self.assertFalse(new_tree.has_both_children())
def test_new_bst_with_value(self):
new_tree = bst(20)
self.assertTrue(new_tree.value != None)
def test_add_node_when_larger(self):
new_tree = bst(20)
larger_node = bst(21)
new_tree.add(larger_node)
self.assertTrue(new_tree.get_right_child().value == larger_node.value)
def test_parent_of_right(self):
new_tree = bst(20)
larger_node = bst(21)
new_tree.add(larger_node)
self.assertTrue(new_tree.get_right_child().get_parent().value == new_tree.value)
prt = prtree()
def printree(node):
dot = prt.dot(node)
dot.view()
prt.dots.append(dot)
# print(dot.source)
# dot.render(
# filename=None, directory=None, view=False, cleanup=False, format='png')
if __name__ == "__main__":
# dots清理等
prt.dots.clear()
t_bst = bst()
t_llrbt = llrbt()
t_sc = scht()
t_lp = lpht()
# 取数据集
d = randoms.dict_int(n=100)
keys = [key for key in d]
minkey = min(keys)
maxkey = max(keys)
# 插入
for k, v in d.items():
t_bst.insert(k, k)
t_llrbt.insert(k, k)
t_sc.insert(k, k)
# test
f.close()
f = open('names_2.txt', 'r')
names_2 = f.read().split("\n") # List containing 10000 names
f.close()
duplicates = [] # Return the list of duplicates in this data structure
# STRETCH STARTS
# set_1 = set(names_1)
# set_2 = set(names_2)
# duplicates = list(set_1.intersection(set_2))
# STRETCH ENDS
# create new tree and set the root value
tree = bst(names_1[0])
# loop over names_1 list and insert names to tree
# so I could utilise bst contains method on it
for name in names_1:
tree.insert(name)
# loop over names_2 list and pass each name to contains method
# of tree in order to find duplicates
# append to duplicates list if found
for name in names_2:
if tree.contains(name):
duplicates.append(name)
end_time = time.time()
print(f"{len(duplicates)} duplicates:\n\n{', '.join(duplicates)}\n\n")
print(f"runtime: {end_time - start_time} seconds")
def test_empty_instantiation(self):
new_tree = bst()
self.assertTrue(new_tree != None)
# bn2_z = bn2(y.data[0,:], model.predictor.b1)
# f.writelines([str(val)+"\n" for val in bn2_z.tolist()])
return z
argvs = sys.argv
unit = 1000
model = L.Classifier(BinaryMLP(784, unit, 10))
if len(argvs) > 1:
chainer.serializers.load_npz(argvs[1], model)
x = np.ones((1, 784), dtype=np.float32) * 128
y1 = forward_linear(model.predictor.l1, x, "tmp/output_y.txt")
z1 = forward_bn(model.predictor.b1, y1, "tmp/output_bn.txt")
h1 = bst(z1)
y2 = forward_linear(model.predictor.l2, h1, "tmp/output_y2.txt")
z2 = forward_bn(model.predictor.b2, y2, "tmp/output_bn2.txt")
h2 = bst(z2)
y3 = forward_linear(model.predictor.l3, h2, "tmp/output_y3.txt")
z3 = forward_bn(model.predictor.b3, y3, "tmp/output_bn3.txt")
#
# chainer.serializers.load_npz(argvs[1], model)
# train, test = chainer.datasets.get_mnist()
# row = train[0]
# data, teacher = row
# model.predictor(data.reshape(1, 784))
#
def __call__(self, x):
h1 = bst.bst(self.b1(self.c1(x), test=not self.train))
h2 = bst.bst(self.b2(self.c2(h1), test=not self.train))
h3 = bst.bst(self.b3(self.c3(h2), test=not self.train))
h4 = bst.bst(self.b4(self.l1(h3), test=not self.train))
return self.b5(self.l2(h4), test=not self.train)
def test_add_node_when_smaller(self):
new_tree = bst(20)
smaller_node = bst(19)
new_tree.add(smaller_node)
self.assertTrue(new_tree.get_left_child().value == smaller_node.value)
def check_tree(tree, lower_bound=None, upper_bound=None):
'''return true if tree is BST'''
if lower_bound and tree.data <= lower_bound:
return False
if upper_bound and tree.data >= upper_bound:
return False
new_upper = min(upper_bound, tree.data) if upper_bound else tree.data
if tree.left and not check_tree(tree.left, lower_bound, new_upper):
return False
new_lower = max(lower_bound, tree.data) if lower_bound else tree.data
if tree.right and not check_tree(tree.right, new_lower, upper_bound):
return False
return True
tree = bst.bst(19)
tree.insert_list([7, 43, 3, 111, 23, 47, 2, 5, 17, 37, 53, 13, 29, 41, 31])
print tree
print check_tree(tree)
tree.left.data = 20
print tree
print check_tree(tree)
tree.left.data = 7
tree.left.right.data = 20
print tree
print check_tree(tree)
start_time = time.time()
f = open('names_1.txt', 'r')
names_1 = f.read().split("\n") # List containing 10000 names
f.close()
f = open('names_2.txt', 'r')
names_2 = f.read().split("\n") # List containing 10000 names
f.close()
duplicates = [] # Return the list of duplicates in this data structure
# Replace the nested for loops below with your improvements
# iterate over list 1
new_bst = bst(names_1[0])
for el in names_1:
new_bst.insert(el)
# see if any elements in names_2 are also in new_bst, if so add them to the duplicates array
for el in names_2:
if new_bst.contains(el):
duplicates.append(el)
end_time = time.time()
print(f"{len(duplicates)} duplicates:\n\n{', '.join(duplicates)}\n\n")
print(f"runtime: {end_time - start_time} seconds")
# ---------- Stretch Goal -----------
# Python has built-in tools that allow for a very efficient approach to this problem
# What's the best time you can accomplish? Thare are no restrictions on techniques or data
