numpy数组拼接 - np.concatenate

将两个numpy数组拼接成一个数组，

import numpy as npa = np.array([[1, 2], [3, 4]])
b = np.array([[5, 6]])# Concatenate along the first axis (rows)
result = np.concatenate((a, b), axis=0)
print(result)

结果如下

[[1 2][3 4][5 6]]

适用场合

一次性计算所有slices向量，可能会撑爆内存，所以多次计算，对计算结果进行拼接

origin:

slices_vecs = sentence_vecs(slices)

print(slices_vecs.shape)

concatenate:

slices_vecs = sentence_vecs(slices[:1])

for i in range(1, len(slices)):
slices_vecs = np.concatenate((slices_vecs, sentence_vecs(slices[i:i+1])), axis=0)

print(slices_vecs.shape)

reference

---