def reverseBetween(self, head, m, n):
"""
:param head: ListNode
:param m: int
:param n: int
:return: ListNode
"""
dummyhead = Node()
dummyhead.next = head
cur = head
index = 1
pre = dummyhead
while index <= n:
if index >= m:
"""todo"""
if index == m:
m_node = cur
m_pre = pre
if index == n:
n_node = cur
n_next = cur.next
next = cur.next
cur.next = pre
pre = cur
cur = next
index += 1
else:
pre = cur
cur = cur.next
index += 1
m_pre.next = n_node
m_node.next = n_next
return dummyhead
def __init__(self, y, y_hat, reg=0.):
"""A node that represents the softmax loss.
Should always be last node in a computational graph.
Only useful for multinomial classification problems.
"""
Node.__init__(self, [y_hat])
self.y = y
self.reg = reg
def reverseList(self, head):
"""
:param head: ListNode
:return: ListNode
"""
dummyhead = Node()
dummyhead.next = head
cur = dummyhead.next
pre = None
while cur != None:
next = cur.next
cur.next = pre
pre = cur
cur = next
dummyhead.next = pre
return dummyhead
def __init__(self, node):
"""Compute the sigmoid of a given input node."""
Node.__init__(self, [node])
def __init__(self, *args):
"""A node that computes the linear combination of a list of input
nodes features, a list of input weights, and a bias term."""
Node.__init__(self, [*args])
def __init__(self):
"""An input node. Input nodes don't perform any computations.
Rather, they represent the input features that will be fed into
the neural network.
"""
Node.__init__(self)
def __init__(self, y, y_hat):
"""A node that computes the mean squared error.
Should only be used at the last node in a network."""
Node.__init__(self, [y, y_hat])
def __init__(self, node, epsilon=1e-4):
"""Computes rectified linear units for the node."""
Node.__init__(self, [node])
self.epsilon = epsilon
def __init__(self, node, epsilon=0., leak=1e-2):
"""Computes leaky rectified linear units for the node."""
Node.__init__(self, [node])
self.epsilon = epsilon
self.leak = leak
def __init__(self, node):
"""A node that represents the softmax activation function.
Should always be last node in a graph before computing loss, and only useful for probabilistic
outputs between 0 and 1."""
Node.__init__(self, [node])