recursion - what's difference between these two statements about passing variables in Python recursively -

i'm new python learner. , there question confusing me lot cause waste time think about.

there algorithm puzzle binary tree, , sum, should find root-to-leaf paths each path's sum equals given sum.

for example: given below binary tree , sum = 22

an example picture here

i have written python recursive method blew , runs correctly on online judgment.

#definition binary tree node. class treenode(object):     def __init__(self, x):         self.val=x         self.left=none         self.right=none class solution(object):     def pathsum(self, root, sum):         """         :type root: treenode         :type sum: int         :rtype: list[list[int]]         """         res=[]         if not root:             return res         temp=[root.val]         self.helper(root,sum,res,temp)         return res     def helper(self, root, sum, res, temp):         if not root:             return 0         if root.left==none , root.right==none , sum==root.val:             res.append(temp)         if root.left!=none:             self.helper(root.left,sum-root.val,res,temp+[root.left.val])         if root.right!=none:             self.helper(root.right,sum-root.val,res,temp+[root.right.val]) 

in last 4 lines, invoke helper function recursively find path sum pass root left child , root right child.

however, if rewrite code below, mean last 4 lines

if root.left!=none:     temp+=[root.left.val]     self.helper(root.left,sum-root.val,res,temp) if root.right!=none:     temp+=[root.right.val]     self.helper(root.right,sum-root.val,res,temp) 

it gives me wrong answer , can't pass online judgment.

dose know difference between 2 kind of ways in pass parameter function in python. or it's declare , pass problem in code.

in view, can't see difference. everyone.help me !

+= alters list in-place:

>>> def inplace(l): ...     l += ['spam'] ... >>> def new_list(l): ...     l = l + ['spam'] ... >>> = ['foo'] >>> inplace(a) >>> ['foo', 'spam'] >>> = ['foo'] >>> new_list(a) >>> ['foo'] 

your original code passes in new list each time:

self.helper(root.left,sum-root.val,res,temp+[root.left.val]) 

but altered code shares temp across recursive calls , extends each time. matters because creating new list gave recursive calls left branch new, independent list right branch of recursion. extending list += give larger list right branch after processing left branch.