Using Memoization to Speed Up Code

# What is memoization?

Yeah, that's right, memo, not memor. I asked my self this when I first saw it. Simply put, memoization can be described as the caching of the results of a sub-step within an algorithm.

That makes it perfect for recursive functions, as they waste plenty of time recomputing everything, so that's what I tried it on, a recursive function.

And what recursive function is better known than a function to calculate the Fibonacci sequence?

After learning about memoization and how to do it, I spent hours perfecting a memoization wrapper which I then used on a JavaScript bigInt Fibonacci function.

I then laughed at the speed differences.

I calculated 1000 iterations of the Fibonacci sequence in 1 second. Without it, it took forever. You can try any number less than 200,000 without choosing to loop, and you'll receive an answer within 10 seconds, guaranteed or I'll refund you.

Try it out yourself!
https://MemoizedFibonacci.studentfires.repl.run

You are viewing a single comment. View All
finlay44111 (77)

what's wrong with something like:

``````int a = 1;
int b = 1;
for (int i = 0; i < 1000; i++)
{
int c = b;
b += a;
a = c;
}``````

that would calculate the 1000th fibonacci number and it would be way faster.

xxpertHacker (627)

@finlay44111 Yeah ik, this practically isn't recursive anymore either. But this is a general wrapper than can be applied to any pure function.

xxpertHacker (627)

@finlay44111 Also, as I told Mr. Economical, this doesn't need to be used on recursive functions, it can be used on any function. It's best to be used on functions that are called often and are very expensive to compute or inefficient at computing.