python - How do I make gmpy array operations faster? -
I am having trouble with speed while trying to use the GMP module. Import NMP as GMP 2 GM = N = 1000 A = Range (N)% Time [for Gmcene (X)) Import as GMP 2 in the 100 # Loop; Most Good 3: 7.3 9 ms per loop% time np esin (a) # 10000 loops, best 3: 198 us per loop
I was thinking that I somehow speed up this calculation Can I give I was thinking that JIT or multi-processing can help, but I do not know how to do it.
Any help would be greatly appreciated if you want to post me more information, please let me know.
I was curious to see how much performance increase would be possible, so gmpy2 , which completely calculates
sin
of a list in C. Unfortunately, there was not much improvement. For the% timeit [gmpy2.sin (x)) 100 loops, the best 3: 4.85 ms per loop% timeit map (gmpy2.sin, a) 100 loops, best 3: 4.59 ms per loop% timeit gmpy2.vector (a) 100 loops, best 3: 4.44 ms per loop
gmpy2
does not issue Global Interpreter Lock (GIL) Threading will not help
Multiprocessing can help, but you will probably need to concretize parts of the code which is in any other process Take seconds (or more) to perform the data to remove the upper part of passing the data.
Software-based, arbitrary-precise floating point is just slow from the original floating point.
Comments
Post a Comment