How can I optimize my NumPy code using Numba?

How can I optimize my NumPy code using Numba?

You can use Numba to compile your NumPy functions to machine code, which can significantly speed up execution. Just decorate your function with @jit.

Can you show me an example of how to use the @jit decorator?

Sure! Here's a simple example:

import numpy as np
from numba import jit

@jit
def sum_array(arr):
    total = 0
    for i in arr:
        total += i
    return total

arr = np.random.rand(1000000)
print(sum_array(arr))

What if I have a function that uses loops and conditionals?

Numba works well with loops and conditionals too! Just ensure that your function is compatible with Numba's supported features.

Are there any limitations I should be aware of when using Numba?

Yes, Numba has some limitations, such as not supporting all Python features and libraries. It's best for numerical computations.