It seems that outside a jitted function, operations on numba.typed.List are very slow compared to the same operation on native Python lists, on my machine the difference is about 20x. In my project I had trouble finding a way to work around the problem, I was wondering if I can get some advice on ho…

Performance of typed.List outside of jit functions

nelson2005 May 12, 2024, 1:54am 3

Like @DannyWeitekamp alluded to, you can build all kinds of fast containers by allocating a numpy array and overlaying control structures on it. One discussion on the topic here

Topic		Replies	Views
Best practices for using read-only Python lists Community Support	4	2044	January 12, 2022
Constructing typed List slow Numba	2	771	April 14, 2022
Numba JIT becoming slower with List Support: How do I do ...?	3	565	April 13, 2022
Typed list of jitted functions in jitclass Support: How do I do ...?	17	3902	November 30, 2023
Why numba decorator cause 10 times slower run Support: How do I do ...?	3	632	August 31, 2022

Performance of typed.List outside of jit functions

Related topics