# Search Algorithm-An Introduction To Binary Search

An algorithm is a set of instructions or steps that are used to solve a problem. In this case, a computer program can be described as a complex algorithm. Normally, algorithms must have a clear beginning and end within a finite amount of steps. And there are steps that allow you to complete an operation. So, what exactly is an algorithm? What is a search algorithm like binary search? To find out answers to these questions, you can also visit algo.monster. It offers professional and informative content about various algorithms.

Binary search is a very common search algorithm. People often use it in many real-life applications. Basically, binary search uses data that is already sorted to make it more effective than sequential search.

**Various kinds of search algorithms**

Searching refers to the process of locating a specific value position among a number of values. It determines whether or not a search target is available in the data. Basically, it is the algorithmic process used to spot a particular item from a group of items.

In the world of computer science, many kinds of search algorithms are available. These useful algorithms include linear search, binary search, interpolation search, Fibonacci search, and many others.

As the fundamental procedure in computing, you can perform search algorithms using an internal data structure or an external one. To a large extent, how fast an application solves problems depends on the search algorithm you choose.

**Binary search: what is it and how does it work?**

The binary search algorithm operates on the principle of divide and conquer due to the fact that it splits the list into halves of each step. As a fast search method using O (log n) for run-time complexity, it is a basic search algorithm in CS. But it only works in sorted data.

Binary search will search for a particular item by looking at the middle item of the list. If your target matches the middle value, the index of the item returns. However, if the value of the mid-item is greater than the target, the item is searched within the sub-array in the lower half. If the middle value is lower, it will keep on the searching process using the other half of the sub-array. And just like that, the process won’t end until the target value is located, or until one is left. Sometimes, the list doesn’t even contain the target you are looking for.

**Binary search: searching process**

This search method is quite different from other search algorithms. Firstly, it begins by looking at the middle item. If the target value is the same as the middle item of the array, return the index. Otherwise, you can then compare the target and middle elements again. So, when this happens, there are two possibilities. The middle item is either higher or lower than the target if it’s not equal, right?

Lower than the target: the search goes on in the higher part of the list.

Higher than the target: it only searches for the target from the items in the lower part. Obviously, it can’t be in the other half.

Thus, as we’ve stated above, the procedure only ends in two ways:

Return the index of the element that was matched when a match has been found. Or else, return -1 when there’s no match at all.

**Binary search offers many benefits**

Unlike linear search in which you have to check each element in the array beginning with the first one.

The best case for both methods is O (1). In the worst case, linear takes N comparisons (n is the length of the list). On the other hand, binary search takes average and worst-case log2(N) comparisons. For instance, the most steps to locate an item in a list of five hundred thousand items, linear search would need 500,000, while binary search needs 20.

For a large number of sorted elements, you can efficiently locate the specific item you want. Generally, it could be the quickest method of finding the target value.

**Binary search has its limitations**

Though efficient mostly, it’s not suitable for small quantities of elements. Instead, you can use linear search.

It works under one condition, elements in the lists are properly sorted. Besides, it should stay sorted if you keep adding new elements to the list. Otherwise, it’s not going to work. You’ll get the wrong result.

It can only be used for element types where there is a less-than relation. Some types are simply not possible to sort (though this is rare).

The list that does not support random access will cause great efficiency losses. It is necessary, for example, to instantly jump to the middle of the list.

Well, as we’ve mentioned, it’s not always the fastest method. Hash lookups could be faster sometimes. Or for unsorted data in small size, linear search could even be better.

**Conclusion **

You can call it binary search/chop, or half-interval search, and logarithmic search. Whatever the name is, it is a type of search that compares the target value with the middle element in a given array contains sorted data.

Also, let’s recap how it works:

Suppose in a list of sorted elements in ascending order, the middle elements is M, the left elements on the left side of M belong to L. Then, the right half of the elements is R. The target element you’re looking for is T.

Spot M. Then, you divide the list into two halves.

If M<T, your search space is R. So, turn right. Eliminate the left half.

If M>T, only search for T in the lower half which is L.

Repeat these 3 steps until you find it or until you find that it’s not in the list.

Binary search performs better than linear with the limitation that it only works if the data is sorted.

For more informative content, you can visit algo.monster. This algorithm expert will give you professional explanations on binary search and other search algorithms as well. Happy learning!