Median of two sorted arrays of different sizes | Set 1 (Linear) - Coding

This is an extension of the median of two sorted arrays of equal size problem. Here we handle arrays of unequal size also.

Examples:

Input: a[] = {1, 12, 15, 26, 38}
       b[] = {2, 13, 24}
Output: 14
Explanation:
After merging arrays the result is
1, 2, 12, 13, 15, 24, 26, 38
median = (13 + 15)/2 = 14.

Input: a[] = {1, 3, 5, 7}
       b[] = {2, 4, 6, 8}
Output: 5
Explanation : 
After merging arrays the result is
1, 2, 3, 4, 5, 6, 7, 8, 9
median = 5

Approach:

The approach discussed in this post is similar to method 1 of equal size median. Use merge procedure of merge sort.
Keep a variable to track of count while comparing elements of two arrays.
Perform merge of two sorted arrays. Increase the value of count as a new element is inserted.
The most efficient way of merging two arrays is to compare the top of both arrays and pop the smaller value and increase the count.
Continue comparing the top/first elements of both arrays until there are no elements left in any arrays
Now to find the median there are two cases to be handled.
- If the sum of length of array sizes is even then store the elements when the count is of value ((n1 + n2)/2)-1 and (n1 + n2)/2 in the merged array add the elements and divide by 2 return the value as median.
- If the value(sum of sizes) is odd then return the (n1 + n2)/2 th element(when value of count is (n1 + n2)/2) as median.

C++

<?php
// A PHP program to find median of 
// two sorted arrays of unequal sizes
 
// A utility function to find 
// median of two integers
/* This function returns median 
of a[] and b[]. Assumptions in this
function: Both a[] and b[] are 
sorted arrays */
function findmedian($a, $n1, $b, $n2)
{
    $i = 0; /* Current index of
               i/p array a[] */
    $j = 0; /* Current index of
               i/p array b[] */
    $k;
    $m1 = -1; $m2 = -1;
    for ($k = 0;
         $k <= ($n1 + $n2) / 2; $k++)
    {
 
        if ($i < $n1 and $j < $n2) 
        {
            if ($a[$i] < $b[$j])
            {
                $m2 = $m1;
                $m1 = $a[$i];
                $i++;
            }
            else
            {
                $m2 = $m1;
                $m1 = $b[$j];
                $j++;
            }
        }
 
        /* Below is to handle the case 
        where all elements of a[] are
        smaller than smallest(or first)
        element of b[] or a[] is empty*/
        else if (i == n1) 
        {
            $m2 = $m1;
            $m1 = $b[j];
            $j++;
        }
 
        /* Below is to handle case
        where all elements of b[] are
        smaller than smallest(or first)
        element of a[] or b[] is empty*/
        else if ($j == $n2) 
        {
            $m2 = $m1;
            $m1 = $a[$i];
            $i++;
        }
    }
 
    /*Below is to handle the case 
    where sum of number of elements 
    of the arrays is even */
    if (($n1 + $n2) % 2 == 0)
        return ($m1 + $m2) * 1.0 / 2;
 
    /* Below is to handle the case 
    where sum of number of elements 
    of the arrays is odd */
    return m1;
}
 
// Driver Code
$a = array( 1, 12, 15, 26, 38 );
$b = array( 2, 13, 24 );
 
$n1 = count($a);
$n2 = count($b);
 
echo(findmedian($a, $n1, $b, $n2));
 
// This code is contributed 
// by inder_verma.
?>

Javascript

<script>
 
// A Javascript program to find median
// of two sorted arrays of unequal sizes
 
// A utility function to find median of two integers
// This function returns median of a[] and b[].
// Assumptions in this function: Both a[] and b[]
// are sorted arrays 
function findmedian(a, n1, b, n2)
{
    let i = 0; /* Current index of
                  i/p array a[] */
    let j = 0; /* Current index of
                  i/p array b[] */
    let k;
    let m1 = -1, m2 = -1;
     
    for(k = 0; k <= (n1 + n2) / 2; k++)
    {
        if (i < n1 && j < n2) 
        {
            if (a[i] < b[j])
            {
                m2 = m1;
                m1 = a[i];
                i++;
            }
            else
            {
                m2 = m1;
                m1 = b[j];
                j++;
            }
        }
         
        /* Below is to handle the case where
        all elements of a[] are
        smaller than smallest(or first)
        element of b[] or a[] is empty*/
        else if (i == n1) 
        {
            m2 = m1;
            m1 = b[j];
            j++;
        }
         
        /* Below is to handle case where
        all elements of b[] are
        smaller than smallest(or first)
        element of a[] or b[] is empty*/
        else if (j == n2) 
        {
            m2 = m1;
            m1 = a[i];
            i++;
        }
    }
     
    /*Below is to handle the case where
    sum of number of elements 
    of the arrays is even */
    if ((n1 + n2) % 2 == 0)
    {
        return (m1 + m2) * 1.0 / 2;
    }
     
    /* Below is to handle the case where
    sum of number of elements 
    of the arrays is odd */
    return m1;
}
 
// Driver code
let a = [ 1, 12, 15, 26, 38 ];
let b = [ 2, 13, 24 ];
let n1 = a.length;
let n2 = b.length;
 
document.write(findmedian(a, n1, b, n2));
 
// This code is contributed by rag2127
 
</script>

Output:

14.000000

Complexity Analysis:

Time Complexity: O(n), Both the lists need to be traversed so the time complexity is O(n).
Auxiliary Space: O(1), no extra space is required.

More Efficient (Log time complexity methods)

Reffered: https://www.geeksforgeeks.org

Searching

Related
Variants of Binary Search
Longest Subarray with first element greater than or equal to Last element
Largest subarray having sum greater than k
Find the only missing number in a sorted array
Ways to choose three points with distance between the most distant points <= L

Type:	Geek
Category:	Coding
Sub Category:	Tutorial
Uploaded by:	Admin
Views:	12