正是如此numpy.repeat http://docs.scipy.org/doc/numpy/reference/generated/numpy.repeat.html does:
>>> a = np.array([0.1, 0.2, 0.3, 0.4, 0.5, 0.6])
>>> s = np.array([3, 3, 9, 3, 6, 3])
>>> np.repeat(a, s)
array([ 0.1, 0.1, 0.1, 0.2, 0.2, 0.2, 0.3, 0.3, 0.3, 0.3, 0.3,
0.3, 0.3, 0.3, 0.3, 0.4, 0.4, 0.4, 0.5, 0.5, 0.5, 0.5,
0.5, 0.5, 0.6, 0.6, 0.6])
在纯 Python 中,你可以执行以下操作:
>>> from itertools import repeat, chain, imap
>>> list(chain.from_iterable(imap(repeat, a, s)))
[0.1, 0.1, 0.1, 0.2, 0.2, 0.2, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.4, 0.4, 0.4, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.6, 0.6, 0.6]
但当然它会比 NumPy 的等价物慢得多:
>>> s = [3, 3, 9, 3, 6, 3]*1000
>>> a = [0.1, 0.2, 0.3, 0.4, 0.5, 0.6]*1000
>>> %timeit list(chain.from_iterable(imap(repeat, a, s)))
1000 loops, best of 3: 1.21 ms per loop
>>> %timeit np.repeat(a_a, s_a) #a_a and s_a are NumPy arrays of same size as a and b
10000 loops, best of 3: 202 µs per loop