I was reading this answer on how to check if a BST is height balanced, and really hooked by the bonus question:
Suppose the tree is massively unbalanced. Like, a million nodes deep on one side and three deep on the other. Is there a scenario in which this algorithm blows the stack? Can you fix the implementation so that it never blows the stack, even when given a massively unbalanced tree?
What would be a good strategy here?
I am thinking to do a level order traversal and track the depth, if a leaf is found and current node depth is bigger than the leaf node depth + 2, then it's not balanced. But how to combine this with height checking?
Edit: below is the implementation in the linked answer
IsHeightBalanced(tree)
return (tree is empty) or
(IsHeightBalanced(tree.left) and
IsHeightBalanced(tree.right) and
abs(Height(tree.left) - Height(tree.right)) <= 1)
To review briefly: a tree is defined as being either null or a root node with pointers .left to a left child and .right to a right child, where each child is in turn a tree, the root node appears in neither child, and no node appears in both children. The depth of a node is the number of pointers that must be followed to reach it from the root node. The height of a tree is -1 if it's null or else the maximum depth of a node that appears in it. A leaf is a node whose children are null.
First let me note the two distinct definitions of "balanced" proposed by answerers of the linked question.
EL-balanced A tree is EL-balanced if and only if, for every node v, |height(v.left) - height(v.right)| <= 1.
This is the balance condition for AVL trees.
DF-balanced A tree is DF-balanced if and only if, for every pair of leaves v, w, we have |depth(v) - depth(w)| <= 1. As DF points out, DF-balance for a node implies DF-balance for all of its descendants.
DF-balance is used for no algorithm known to me, though the balance condition for binary heaps is very similar, requiring additionally that the deeper leaves be as far left as possible.
I'm going to outline three approaches to testing balance.
Size bounds for balanced trees
Expand the recursive function to have an extra parameter, maxDepth. For each recursive call, pass maxDepth - 1, so that maxDepth roughly tracks how much stack space is left. If maxDepth reaches 0, report the tree as unbalanced (e.g., by returning "infinity" for the height), since no balanced tree that fits in main memory could possibly be that tall.
This approach relies on an a priori size bound on main memory, which is available in practice if not in all theoretical models, and the fact that no subtrees are shared. (PROTIP: unless you're very careful, your subtrees will be shared at some point during development.) We also need height bounds on balanced trees of at most a given size.
EL-balanced Via mutual induction, we prove a lower bound, L(h), on the number of nodes belonging to an EL-balanced tree of a given height h.
The base cases are
L(-1) = 0
L(0) = 1,
more or less by definition. The inductive case is trickier. An EL-balanced tree of height h > 0 is a node with an EL-balanced child of height h - 1 and another EL-balanced child of height either h - 1 or h - 2. This means that
L(h) = 1 + L(h - 1) + min(L(h - 2), L(h - 1)).
Add 1 to both sides and rearrange.
L(h) + 1 = L(h - 1) + 1 + min(L(h - 2) + 1, L(h - 1) + 1).
A little while later (spoiler), we find that
L(h) <= phi^(h + 2)/sqrt(5),
where phi = (1 + sqrt(5))/2 ~ 1.618.
maxDepth then should be set to the floor of the base-phi logarithm of the maximum number of nodes, plus a small constant that depends on fenceposty things.
DF-balanced Rather than write out an induction proof, I'm going to appeal to your intuition that the worst case is a complete binary tree with one extra leaf on the bottom. Then the proper setting for maxDepth is the base-2 logarithm of the maximum number of nodes, plus a small constant.
Iterative deepening depth-first search
This is the theoretician's version of the answer above. Because, for some reason, we don't know how much RAM our computer has (and with logarithmic space usage, it's not as though we need a tight bound), we again include the maxDepth parameter, but this time, we use it to truncate the tree implicitly below the specified depth. If the height of the tree comes back below the bound, then we know that the algorithm ran successfully. Alternatively, if the truncated tree is unbalanced, then so is the whole tree. The problem case is when the truncated tree is balanced but with height equal to maxDepth. Then we increase maxDepth and retry.
The simplest retry strategy is to increase maxDepth by 1 every time. Since balanced trees with n nodes have height O(log n), the running time is O(n log n). In fact, for DF-balanced trees, the running time is also O(n), since, except for the last couple traversals, the size of the truncated tree increases by a factor of 2 each time, leading to a geometric series.
Another strategy, doubling maxDepth each time, gives an O(n) running time for EL-balanced trees, since the largest tree of height h, with 2^(h + 1) - 1 nodes, is much smaller than the smallest tree of height 2h, with approximately (phi^2)^h nodes. The downside of doubling is that we may use twice as much stack space. With increase-by-1, however, in the family of minimum-size EL-balanced trees we constructed implicitly in defining L(h), the number of nodes at depth h - k in the tree of height h is polynomial of degree k. Accordingly, the last few scans will incur some superlinear term.
Temporarily mutating pointers
If there are parent pointers, then it's easy to traverse depth-first in place, because the parent pointers can be used to derive the relevant information on the stack in an efficient manner. If we don't have parent pointers but can mutate the tree temporarily, then, for descent into a child, we can cannibalize the pointer to that child to store temporarily the node's parent. The problem is determining on the way up whether we came from a left or a right child. If we can sneak a bit (say because pointers are 2-byte aligned, or because there's a spare bit in the balance factor, or because we're copying the tree for stop-and-copy garbage collection and can determine which arena we're in), then that's one way. Another test assumes that the tree is a binary search tree. It turns out that we don't need additional assumptions, however: Explain Morris inorder tree traversal without using stacks or recursion .
The one fly in the ointment is that this approach only works, as far as I know, on DF-balance, since there's no space on the stack to put the partial results for EL-balance.
Related
I'm trying to figure out this data structure, but I don't understand how can we
tell there are O(log(n)) subtrees that represents the answer to a query?
Here is a picture for illustration:
Thanks!
If we make the assumption that the above is a purely functional binary tree [wiki], so where the nodes are immutable, then we can make a "copy" of this tree such that only elements with a value larger than x1 and lower than x2 are in the tree.
Let us start with a very simple case to illustrate the point. Imagine that we simply do not have any bounds, than we can simply return the entire tree. So instead of constructing a new tree, we return a reference to the root of the tree. So we can, without any bounds return a tree in O(1), given that tree is not edited (at least not as long as we use the subtree).
The above case is of course quite simple. We simply make a "copy" (not really a copy since the data is immutable, we can just return the tree) of the entire tree. So let us aim to solve a more complex problem: we want to construct a tree that contains all elements larger than a threshold x1. Basically we can define a recursive algorithm for that:
the cutted version of None (or whatever represents a null reference, or a reference to an empty tree) is None;
if the node has a value is smaller than the threshold, we return a "cutted" version of the right subtree; and
if the node has a value greater than the threshold, we return an inode that has the same right subtree, and as left subchild the cutted version of the left subchild.
So in pseudo-code it looks like:
def treelarger(some_node, min):
if some_tree is None:
return None
if some_node.value > min:
return Node(treelarger(some_node.left, min), some_node.value, some_node.right)
else:
return treelarger(some_node.right, min)
This algorithm thus runs in O(h) with h the height of the tree, since for each case (except the first one), we recurse to one (not both) of the children, and it ends in case we have a node without children (or at least does not has a subtree in the direction we need to cut the subtree).
We thus do not make a complete copy of the tree. We reuse a lot of nodes in the old tree. We only construct a new "surface" but most of the "volume" is part of the old binary tree. Although the tree itself contains O(n) nodes, we construct, at most, O(h) new nodes. We can optimize the above such that, given the cutted version of one of the subtrees is the same, we do not create a new node. But that does not even matter much in terms of time complexity: we generate at most O(h) new nodes, and the total number of nodes is either less than the original number, or the same.
In case of a complete tree, the height of the tree h scales with O(log n), and thus this algorithm will run in O(log n).
Then how can we generate a tree with elements between two thresholds? We can easily rewrite the above into an algorithm treesmaller that generates a subtree that contains all elements that are smaller:
def treesmaller(some_node, max):
if some_tree is None:
return None
if some_node.value < min:
return Node(some_node.left, some_node.value, treesmaller(some_node.right, max))
else:
return treesmaller(some_node.left, max)
so roughly speaking there are two differences:
we change the condition from some_node.value > min to some_node.value < max; and
we recurse on the right subchild in case the condition holds, and on the left if it does not hold.
Now the conclusions we draw from the previous algorithm are also conclusions that can be applied to this algorithm, since again it only introduces O(h) new nodes, and the total number of nodes can only decrease.
Although we can construct an algorithm that takes the two thresholds concurrently into account, we can simply reuse the above algorithms to construct a subtree containing only elements within range: we first pass the tree to the treelarger function, and then that result through a treesmaller (or vice versa).
Since in both algorithms, we introduce O(h) new nodes, and the height of the tree can not increase, we thus construct at most O(2 h) and thus O(h) new nodes.
Given the original tree was a complete tree, then it thus holds that we create O(log n) new nodes.
Consider the search for the two endpoints of the range. This search will continue until finding the lowest common ancestor of the two leaf nodes that span your interval. At that point, the search branches with one part zigging left and one part zagging right. For now, let's just focus on the part of the query that branches to the left, since the logic is the same but reversed for the right branch.
In this search, it helps to think of each node as not representing a single point, but rather a range of points. The general procedure, then, is the following:
If the query range fully subsumes the range represented by this node, stop searching in x and begin searching the y-subtree of this node.
If the query range is purely in range represented by the right subtree of this node, continue the x search to the right and don't investigate the y-subtree.
If the query range overlaps the left subtree's range, then it must fully subsume the right subtree's range. So process the right subtree's y-subtree, then recursively explore the x-subtree to the left.
In all cases, we add at most one y-subtree in for consideration and then recursively continue exploring the x-subtree in only one direction. This means that we essentially trace out a path down the x-tree, adding in at most one y-subtree per step. Since the tree has height O(log n), the overall number of y-subtrees visited this way is O(log n). And then, including the number of y-subtrees visited in the case where we branched right at the top, we get another O(log n) subtrees for a total of O(log n) total subtrees to search.
Hope this helps!
I need to write an algorithm (in pseudo-code) that chech if a given BST tree is a valid AVL tree. In doing that I need to give to each node a rank (in AVL trees rank means the height of the node) so the outcome will be a valid AVL tree.
I thought about a simple algorithm that calculates in each step the height of a node and the height of its two sons (if the sons is null then the height is -1), and then checks if the difference between the heights is 1,1 or 1,2 or 2,1. If not then it's not an AVL tree. If yes than we do the same for node.left and node.right.
My problem with that algorithm is that the time complexity is very huge and could go even to O(n^2). Is there a more efficient algorithm?
Another algorithm that I want to find is when given a valid AVL tree and the rank of each node (rank=height), I need to find a series of inserts that build the same tree. I thought about doing it by sorted order of the keys, but the outcome isn't right.
AVL-Tree check
You actually got the right idea. But you missed out on using the recursive definition of the height of a tree
height(node) = 1 + max(height(node.left), height(node.right))
which is why you got an enormous complexity (though O(2^n) is too pessimistic). Instead of directly calculating the height of a node in a top-bottom-approach, you could go the other way and calculate the height of the individual nodes bottom-top:
valid_avl(node):
if node is null then
return -1, True
left_height, left_valid = valid_avl(node.left)
right_height, right_valid = valid_avl(node.right)
if not left_valid or not right_valid or abs(left_height - right_height) > 1 then
return -1, False
else
return 1 + max(left_height, right_height), True
You might want to split the this function into two functions and avoid the use of tuples, depending on the language you use. Note that the above is albeit looking similar to python just pseudocode!!!
Rebuilding the tree
This ones actually fairly simple. Get all values in the tree in level-order and insert them exactly in this order. This way the tree never gets rebalanced and each node is automatically placed in the correct position.
I'm looking for a efficient way of merging two BST, knowing that the elements in the first tree are all lower than those in the second one.
I saw some merging methods but without that feature, i think this should simplify the algorithm.
Any idea ?
If the trees are not balanced, or the result shouldn't be balanced that's quite easy:
without loss of generality - let the first BST be smaller (in size) than the second.
1. Find the highest element in the first BST - this is done by following the right son while it is still available. Let the value be x, and the node be v
2. Remove this item (v) from the first tree
3. Create a new Root with value x, let this new root be r
4. set r.left = tree1.root, r.right = tree2.root
(*) If the first BST is bigger in size than the second, just repeat the process for finding v as the smallest node in the second tree.
(*) Complexity is O(min{|T1|,|T2|}) worst case (finding highest element if the tree is very inbalanced), and O(log(min{|T1|,|T2|})) average case - if the tree is relatively balanced.
I've just learn B-tree and B+-tree in DBMS.
I don't understand why a non-leaf node in tree has between [n/2] and n children, when n is fix for particular tree.
Why is that? and advantage of that?
Thanks !
This is the feature that makes the B+ and B-tree balanced, and due to it, we can easily compute the complexity of ops on the tree and bound it to O(logn) [where n is the number of elements in the data set].
If a node could have more then B sons, we could create a tree with depth 2: a root, and all other nodes will be leaves, from the root. searching for an element will be then O(n), and not the desired O(logn).
If a node could have less then B/2 sons, we could create a tree which is actually a linked list [n nodes, each with 1 son], with height n - and a search op will again be O(n) instead of O(logn)
Small currection: every non-leaf node - except the root, has B/2 to B children. the root alone is allowed to have less then B/2 sons.
The basic assumption of this structure is to have a fixed block size, this is why each internal block has n slots for indexing its children.
When there is a need to add a child to a block that is full (has exactly n children), the block is split into two blocks, which then replace the original block in its parent's index. The number of children in each of the two blocks is obviously n div 2 (assuming n is even). This is what the lower limit comes from.
If the parent is full, the operation repeats, potentially up to the root itself.
The split operation and allowing for n/2-filled blocks allows for most of the insertions/deletions to only cause local changes instead of re-balancing huge parts of the tree.
I had once known of a way to use logarithms to move from one leaf of a tree to the next "in-order" leaf of a tree. I think it involved taking a position value (rank?) of the "current" leaf and using it as a seed for a fresh traversal from the root down to the new target leaf - all the way using a log function test to determine whether to follow the right or left node down to the leaf.
I no longer recall how to exercise that technique. Can anyone re-introduce me?
I also don't recall if the technique required the tree to be balanced, or if it worked on n-trees or only binary trees. Any info would be appreciated.
Since you mentioned whether to go left or right, I'm going to assume you're talking about a binary tree specifically. In that case, I think you're right that there is a way. If your nodes are numbered left-to-right, top-to-bottom, starting with 1, then you can find the rank (depth in the tree) by taking the log2 of the node's number. To find that node again from the root, you can use the binary representation of the number, where 0 = left and 1 = right.
For example:
n = 11
11 in binary is 1011
We always ignore the first 1 since it's going to be there for every number (all nodes of rank n will be binary numbers with n+1 digits, with the first digit being 1). We're left with 011, which is saying from the root go left, then right, then right.
If you want to find the next in-order leaf, take the current leaf's number and add one, then traverse from the root using this method.
I believe this only works with balanced binary trees.
OK, this proposal requires more characters than I can fit into a comment box. Steven does not believe that knowing the depth of the node in the tree is useful. I think it is. I have been wrong in the past, and I'm sure I'll be wrong in the future, so I will try to explain how this idea works in an attempt to not be wrong in the present. If I am, I apologize ahead of time. I'm nearly certain I got it from one of my Algorithms and Datastructures courses, using the CLR book. Please excuse any slips in notation or nomenclature, I haven't studied this stuff in a while.
Quoting wikipedia, "a complete binary tree is a binary tree in which every level, except possibly the last, is completely filled, and all nodes are as far left as possible."
We are considering a complete tree with any branching degree (where a binary tree has a branching degree of two). Also, we are considering our nodes to have a 'positional value' which is an ordering of the positional value (top to bottom, left to right) of the node.
Now, if we are given a positional value, we can find the node in the following fashion. Take the log_base_n of the positional value of the element we are looking for (floor of this, we want an integer). Traverse down from the root that many times, minus one. Now, start looking through all the children of the nodes at this level. Your node you are searching for will be in this set.
This is an attempt in explaining the additional part of the wikipedia definition:
"This depth is equal to the integer part of log2(n) where n
is the number of nodes on the balanced tree.
Example 1: balanced tree with 1 node, log2(1) = 0 (depth = 0).
Example 2: balanced tree with 3 nodes, log2(3) = 1.59 (depth=1).
Example 3: balanced tree with 5 nodes, log2(5) = 2.32
(depth of tree is 2 nodes)."
This is useful, because you can simply traverse down to this level and then start looking around. It is useful and important to know the depth your node is located on, so you can start looking there, instead of starting to look at the beginning. Unless you know what level of the tree you are on, you get to start looking at all the nodes sequentially.
That is why I think it is helpful to know the depth of the node we are searching for.
It is a little bit odd, since having the "positional value" is not something we normally care about in a tree. I can see why Steve thought of this in terms of an array, since positional value is inherent in arrays.
-Brian J. Stinar-
Something that at least resembles your description is the Binary Heap, used a.o. in Priority Queues.
I think I've found the answer, or at least a facsimile.
Assume the tree nodes are numbered, starting at 1, top-down and left-to-right. Assume traversal begins at the root, and halts when it finds node X (which means the parent is linked to its children). Also, for quick reference, the base 2 logarithmic values for nodes 1 through 12 are:
log2(1) = 0.0
log2(2) = 1
log2(3) = 1.58
log2(4) = 2
log2(5) = 2.32
log2(6) = 2.58
log2(7) = 2.807
log2(8) = 3
log2(9) = 3.16
log2(10) = 3.32
log2(11) = 3.459
log2(12) = 3.58
The fractional portion represents a unique diagonal position (notice how nodes 3, 6, and 12 all have fractional portion 0.58). Also notice that every node belongs either to the left or right side of the tree, depending on whether the log fractional component is less or great than 0.5. Anecdotes aside, the algorithm for finding a node is then as follows:
examine fractional portion, if it is less than .5, turn left. Else turn right.
subtract one from the whole number portion of the log, stop if the value reaches zero.
double the fractional portion, and start over.
So, for example, if node 11 is what you seek then you start by computing the log which is 3.459. Then...
3-459 <=fraction less than .5: turn left and decrement whole number to 2.
2-918 <=doubled fraction more than .5: turn right and decrement whole number to 1.
1-836 <=doubling .918 gives 1.836: but only fractional part counts: turn right and dec prior whole number to 0. Done!!
With appropriate accomodations, the same technique appears to work for any balanced n-ary tree. For example, given a balanced ternary tree, the choice of following left, middle, or right edges is again based on the fractional portion of the log, as follows:
between 0.5-0.832: turn left (a one-third fraction range)
between 0.17-0.49: turn right (another one-third fraction range)
otherwise go down the middle. (the last one-third range)
The algorithm is adjusted by multiplying the fractional portion by 3 instead of 2. Again, a quick reference for those who want to test this last statement:
log3(1) = 0.0
log3(2) = 0.63
log3(3) = 1
log3(4) = 1.26
log3(5) = 1.46
log3(6) = 1.63
log3(7) = 1.77
log3(8) = 1.89
log3(9) = 2
At this point I wonder if there is an even more concise way to express this whole "log-based top-down selection of a node." I'm interested if anyone knows...
Case 1: Nodes have pointers to their parent
Starting from the node, traverse up the parent pointer until one with non-null right_child is found. Go to the right_child and traverse left_child as long as they are non-null.
Case 2: Nodes do not have pointers to the parent
Starting from the root, find the path to the node (including the root and the node). Then find the latest vertex (i.e. a node) in the path that has non-null right_child. Go the the right_child and traverse left_child as long as they are non-null.
In both cases, we traversing either up or down from the root to one of the nodes. The maximum of such traversal is in the order of the depth of the tree, hence logarithmic in the size of the nodes if the tree is balanced.