Torchquad.py

Go to the documentation of this file.

 
 
import torch
import numpy as np
from torchquad import MonteCarlo, set_up_backend
 
 
set_up_backend('torch', data_type='float32')
 
 
def integrate(f, num_dim: int, num_iter: int, num_warmup: int, num_calls: int, limits: list[tuple[float]]=[]):
    limits = limits if len(limits) > 0 else num_dim * [(0., 1.)]
 
    def func(xarr):
        return torch.from_numpy(np.array([f([float(x) for x in xvals]) for xvals in xarr.numpy()]))
 
    mc = MonteCarlo()
    res = mc.integrate(func, dim=num_dim, N=num_calls, integration_domain=limits, backend='torch')
    return (float(res.numpy()), 1.)
 
 
if __name__ == '__main__':
    import math
    print(integrate(lambda x: x[0]**2 + x[1]**2, 2, 10, 1000, 1000))
    print(integrate(lambda x: math.sin(x[0]), 1, 10, 1000, 1000, [(0, math.pi)]))