17 ms·
For numpy, it's correct. But numba can optimize the loop. So optimized and readable loop is not an advantage of julia compare to numba.
by longqzh 5y ago
For numpy, it's correct. But numba can optimize the loop. So optimized and readable loop is not an advantage of julia compare to numba.
- ChrisRackauckas 5y agoNot in real-world contexts. This is spelled out in Julia for Biologists (https://arxiv.org/abs/2109.09973 https://arxiv.org/abs/2109.09973) which does the operation counting to show why using Numba with SciPy is still an order of magnitude slower in scientific operations like solving differential equations compared to Julia. An order of magnitude on widely used scientific analyses is pretty significant!