# Regularity condition in the master theorem.

Master theorem is a direct way to get the solution of a recurrence relation, provided that it is of the following type:

T(n) = aT(n/b) + f(n) where a >= 1 and b > 1

The theorem consists of the following three cases:

**1**.If f(n) = () where c < then T(n) = (n)

**2**.If f(n) = () where c = then T(n) = (Log n)

**3**.If f(n) = () where c > then T(n) = (f(n))

Imagine the recurrence **aT(n/b) + f(n)** in the form of a tree.

**Case 1** covers the case when the children nodes does more work than the parent node.

For example the equation **T(n)=2T(n/2)+1** falls under the category of case 1 and we can clearly see from it’s tree below that at each level the children nodes perform twice as much work as the parent node.

T(n) ------(1) / \ T(n/2) T(n/2) ------(2) / \ / \

**Case 2** covers the case when the children node and the parent node does an equal amount of work.

For example the equation **T(n)=2T(n/2)+n** falls under the category of case 2 and we can clearly see from it’s tree below that at each level the children nodes perform as much work as the parent node.

T(n) ------(n) / \ T(n/2) T(n/2) ------(n) / \ / \

**Case 3** covers the scenario that the parent node does more work than the children nodes.

**T(n)=T(n/2)+n** is a example of case 3 where the parent performs more work than the child.

T(n) ------(n) | T(n/2) ------(n/2) |

In case 1 and case 2 the case conditions themselves make sure that work done by children is either more or equal to the parent but that is not the case with case 3.

In case 3 we apply a **regulatory condition** to make sure that the parent does at least as much as the children.

The regulatory condition for case 3 is

af(n/b)<=cf(n).

This says that f(n) (the amount of work done in the root) needs to be at least as big as the sum of the work done in the lower levels.

The equation **T(n) = T(n/2) + n(sin(n – /2) + 2)** is a example where regulatory condition makes a huge difference. The equation isn’t satisfying case 1 and case 2. In case 3 for large values of n it can never satisfy the regulatory condition. Hence this equation is beyond the scope of master theorem.

This article is contributed by **Vineet Joshi**. If you like GeeksforGeeks and would like to contribute, you can also write an article using contribute.geeksforgeeks.org or mail your article to contribute@geeksforgeeks.org. See your article appearing on the GeeksforGeeks main page and help other Geeks.

Please write comments if you find anything incorrect, or you want to share more information about the topic discussed above.

GeeksforGeeks has prepared a complete interview preparation course with premium videos, theory, practice problems, TA support and many more features. Please refer Placement 100 for details

## Recommended Posts:

- Master Theorem For Subtract and Conquer Recurrences
- Advanced master theorem for divide and conquer recurrences
- Paranthesis Theorem
- Count of pairs satisfying the given condition
- Partition the digits of an integer such that it satisfies a given condition
- Loop Invariant Condition with Examples of Sorting Algorithms
- Count of subarrays of size K which is a permutation of numbers from 1 to K
- Create an array such that XOR of subarrays of length K is X
- XOR of elements in a given range with updates using Fenwick Tree
- Minimum valued node having maximum depth in an N-ary Tree
- Count of elements which are power of 2 in a given range subarray for Q queries
- Two player game in which a player can remove all occurrences of a number
- Maximize 3rd element sum in quadruplet sets formed from given Array
- Count of Double Prime numbers in a given range L to R