· 4 Min read

Write FAST Python using Automatic Memoization ⚡

I came across this decorator in Functools which enables automatic Memoization of a function (Basically, storing what a certain function call returns and not having to run it again)

I'm about to blow your minds. and make the code more than 1000 times faster with just ONE line of code!

So here's a snippet that takes about 52 seconds to run. It's very slow and bad code because it's having to run the same fib() function multiple times. So every function call takes twice the amount of time the last function call. oh what a mess bad slow code

Here's the magic ✨

By using the functool library's cache decorator, you can memoise the code automatically for much faster results! Literally just one line of code (and the import, obviously)

from functools import cache
 
@cache
def fib(n):
    ...

Just adding this makes the code run in JUST 0.013621500000000009 seconds! What's the cache here?(ha, get it? Catch ? cache?) Basically all that the decorator is doing, is "remembering" the function call in memory, so this takes more RAM. But there's a way to make this even better.

It's by using the recently-used-cache decorator (lru_cache)

from functools import lru_cache
 
@lru_cache(maxsize=5)

This only stores the last 5 function calls, which is much better. So here's the complete code:

LRU cache

How mind blowing is that!!

If you liked this little python trick, make sure to 💖 this post. If you REALLY liked it, Follow me here - https://dev.to/dhravya and if you REALLY REALLY liked it, follow me on twitter! https://twitter.com/dhravyashah