Thread: Generating the Mode of a Data Set

Hello,
I'm working with a large data set and was wondering if there were some statistical shortcuts. I'm trying to compute the mode of a data set and can only think of an O(n^2) solution. Ordinarily, this wouldn't be too bad, but I feel like there might be something else that I could do. This is what I was thinking of trying:

Code:

int count=0,temp,mode=nums[0];
for (i=0;i<setMax;i++)
    {
     temp =0;
     for(j=0;j<setMax;j++)
         if(nums[j]==nums[i]) temp++;
     if(temp>count)
        {
         temp=count;
         mode = nums[i];
         };
     };

When I used the mode function in excel, there was a noticeable amount of time in between when I pressed enter and when the function returned something around 1.7 (approx .25sec). Since I have to compute the mode many times, I hope that I can optimize it without having to do something too complicated. Obviously, I could used a linked list with frequencies, but then other parts of my code would slow down unreasonably.

Thanks for you help.

If your numbers are integers, you can do bucket sort (we just had this discussion, I think).

bucketsort was mentioned below:

http://cboard.cprogramming.com/showt...ght=bucketsort

Just for curiosity. Thought this was cool

http://vision.bc.edu/~dmartin/teachi...-html/all.html