python - 使用GPU加速python中的蒙特卡洛仿真(Using GPU Speeding up monte carlo simulation in python)

Question

The code I have uses monte carlo techniques to plot points where an electron would be found around a hydrogen atom but when it comes to plotting this the program can take upwards of a few hours to produce the graph I've been looking at ways to reduce this time and have tried using numba I don't know if I'm using it correctly but it hasn't made any difference

(我使用的代码使用蒙特卡洛技术绘制出在氢原子周围发现电子的点，但是当涉及到绘制该程序时，该程序可能要花费几个小时才能生成该图，我一直在研究如何减少该图这次并尝试使用numba，我不知道我是否正确使用它，但没有任何区别)

can someone please help me

(有人可以帮帮我吗)

from scipy.special import genlaguerre, sph_harm
from numpy import random, linspace, sqrt, pi, arccos, exp
from matplotlib.pyplot import plot, figure, show
from numba import jit

n = 2
l = 1
m = -1

Lx = 3*10**-9
Ly = Lx

a_0 = 5.29*10**-11

x = linspace(0,Lx, 100000)
fi = linspace(0,2*pi, 100000)


def fact(x):
    if x == 0:
        return 1
    else:
        return x*fact(x-1)

def row(r):
    return (2*r)/(n*a_0)

def Hydrograd(r):
    return -1*sqrt((2/(n*a_0))**3*(fact(n - l- 1))/(2*n*(fact(n+l))**3))*exp(-1*row(r)/2)*row(r)**l*genlaguerre(n-l-1, 2*l+1)(row(r))

def Hydrogsph(theta, phi):
    return sph_harm(m,l,theta,phi).real

@jit
def HydroFunc(r,theta, phi):
    return abs(Hydrograd(r))**2*abs(Hydrogsph(theta, phi))**2

def rFinder(x,y,z):
    return sqrt(x**2 + y**2 + z**2)

def phiFinder(x,y,z):
    return arccos(x/rFinder(x,y,z))

i = 0
p = 50
@jit
def monty(Lx,Ly,x,fi):
    i = 0
    p = 50
    X = []
    Y = []
    while i < p:
        xr = random.uniform(-Lx,Lx)
        yr = random.uniform(-Ly,Ly)
        if HydroFunc(rFinder(xr,yr,0),0,phiFinder(xr,yr,0))>random.uniform(0,max(HydroFunc(x,0,fi))/100):
            X.append(xr)
            Y.append(yr)
            i += 1
            print(i)
    figure()
    plot(X,Y,'.')
    show()

monty(Lx,Ly,x,fi)

  ask by Jack Weaver translate from so

Answer 1

等待大神答复