通过分段函数组装numpy数组

Question

我正在寻找通过间隔函数组装numpy数组的快速方法。这是我现在的代码（简化），它耗费了大量时间来创建子数组：

import numpy as np

def func():
    a = np.random.random(10000) # avarage size of my arrays
    b = np.random.random(10000)
    r = np.zeros_like(a)

    ind = np.where(a > 0.1)[0]  # some data close to limits is always cut
    ind = ind[a[ind] < 0.9]

    ind2 = ind[a[ind] <= 0.3]  # first interval
    r[ind2] = a[ind2] * b[ind2]  # exemplary calculation

    ind = ind[ind != ind2][0]

    ind2 = ind[a[ind] <= 0.8]  # second interval
    r[ind2] = a[ind2] * b[ind2] * 5

    ind = ind[ind != ind2][0]

    ind2 = ind[a[ind] > 0.8]  # even up to 20 intervals (can be done in loop, 
                                        # here shown separated for clarity)
    r[ind2] = a[ind2] * b[ind2] * 10

    return r

我会感激每一个提示。