2,016 views

We wish to construct a $B^+$ tree with fan-out (the number of pointers per node) equal to $3$ for the following set of key values:

$80, 50, 10, 70, 30, 100, 90$

Assume that the tree is initially empty and the values are added in the order given.

1. Show the tree after insertion of $10$, after insertion of $30$, and after insertion of $90$. Intermediate trees need not be shown.
2. The key values $30$ and $10$ are now deleted from the tree in that order show the tree after each deletion.

fanout means no of records per node but b+ trees contain an additional pointer pointing to next node so in this case would splitting will take place after insertion of 2 key values or 3 key values? Plz explain...
After 2 key values. Also, that additional pointer appears only in leaves.
why split after 2 key value ??

since order = 3, so we can have 2 key elements in a node with no problem but when we have 3 key then it should be a overflow right ??
plz explain this one....
Same like btree
B+ tree with fan-out?

What does that mean.

Anyway its always fan-out only.

### Subscribe to GO Classes for GATE CSE 2022

$(a) \ B^{+}$ tree insertion: $80,50,10,70,30,100,90$

Order of $B^{+} = p =3$

• Overflow: When number of key values exceed $p-1=3-1 = 2$

Tree after insertion of  $10:$

Tree after insertion of  $30:$

Tree after insertion of  $90:$

$(b)$

• Underflow: if leaf node contain $\left \lceil \frac{p}{2}\right \rceil - 1 = 2-1 = 1$ key values.

Deletion of key-value $30:$

Deletion of key-value $10:$

Here when we delete the key-value $10$ then underflow happened, so we can merge this node with the right sibling.

When we merge two nodes then the parent node one value needs to come down i.e. $50$ here.

Now if we try to bring $50$ down then that node will again suffer from underflow as it will become empty.

so we will try to merge this node it with its right sibling i.e. the node which contains $70$

Again, When we merge two nodes(nodes in the 2nd level that contain 50 and 70) then one value of the parent node (i.e. node having $50,70$) needs to come down

So we will bring $70$ down and merge it with $50$ since bringing $70$ down will not cause underflow as $80$ is present in the parent node.

@Lakshman Patel RJIT

Did you write all the steps?

Because I am getting confused reading this.

As per my understanding in case of internal node overflow in B+ Tree the Smallest value among the keys values on Right hand of the median is not copied to the internal node as happens in case of B-Tree also ,

so after inserting 100 , 70 will become the root node but it’s value won’t be copied to the internal node below the root node.

in the B+ tree order of internal node and order of leaf node are different

order of internal node is maximum no of children

order of leaf node is maximum no of key-value pair in the leaf

in the question, it is given that

B+ tree with fan-out (the number of pointers per node) equal to 3

so from this, we can easily conclude that the order of the internal node is 3

and order of leaf node is 2   ( just think about it why it is 2 ?? you will get the answer )

for internal node

since order = 3, so we can have 2 key elements in a node this is  not a problem but when we have 3 key then we have to split them

for leaf node

since order = 2, so we can have 2 key elements in a node with no problem but when we have 3 key then we have to split them

correction in the diagram: every leaf node will point to the next leaf node, by doing this we can observe the fanout of each node is 3

Q: why the order of leaf node is 2?

it is because given fan-out (the number of pointers per node) equal to 3

one pointer point to next leaf and two-pointer point to the record means there are maximum  two record pointer

and order of leaf node is the maximum number of key-value pair

so it is 2

### 1 comment

Few doubts:

1. After inserting 100 and splitting the nodes, the node containing 70,80 is pointing to 2 child nodes. Is it correct. Which  pointer will be NULL?

2. In final tree, Root node is having 2 elements. But in Root node there should be only one element. So to split this node so that root contains one element

While splitting an internal node we push the middle element to parent and we don’t maintain a copy of that in child internal node.

But while spitting a leaf node we push the middle one to parent and we keep a copy in leaf node as well.

a)

We don’t keep a copy of 70 in child internal  node when inserting 100

b)After deleting 30 and 10 relsuting tree is

Play with this tool for better understanding

https://www.cs.usfca.edu/~galles/visualization/BPlusTree.html