使用 np.tile 生成 (n, 1, 2) 数组

Question

I want to create n times (1,2) arrays and each array should have the same elements. First I generate n times 1 D array and then I use a loop to iterate over these elements and repeat each element to fill (n, 1,2) array. my code is the following:

import numpy as np


def u_vec():
   return np.array([np.random.rand(1)])

n=10
u1 = np.zeros(n)

for i in range(n):
    u1[i] = u_vec()

print(u1)

def u_vec1():
    u_vec = np.zeros((n, 2,1))
    for i in range(len(u1)):
        u_vec[i] += np.tile(u1[i], (2,1))
    return u_vec

u = u_vec1()
print(u)

the output that I get is

[0.4594466  0.80924903 0.3186138  0.03601917 0.9116031  0.68199505
 0.78999837 0.33778259 0.97626521 0.84925156]


[[[0.4594466 0.4594466]]

[[0.        0.       ]]

[[0.        0.       ]]

[[0.        0.       ]]

[[0.        0.       ]]

[[0.        0.       ]]

[[0.        0.       ]]

[[0.        0.       ]]

[[0.        0.       ]]

[[0.        0.       ]]]

I do not understand why only the first element is filled but the others are filled with zero. Could someone please help me? Thank you very much! the output that I would like to have

[[[0.4594466 0.4594466]]

[[0.3186138  0.3186138]]

[[ 0.03601917 0.03601917]]

[[ 0.9116031  0.9116031 ]]

[[0.68199505  0.68199505]]

[[0.78999837  0.78999837]]

[[0.33778259  0.33778259]]

[[0.97626521   0.97626521]]

[[0.84925156   0.84925156]]]]

Answer 1

I see the problem. The problem is that your return u_vec statement is enclosed in the for loop. So only the first subarray is updated with the random values and the rest of u_vec remains 0 because you return immediately after the first iteration of the for loop. You should use

def u_vec1():
    u_vec = np.zeros((n, 2,1))
    for i in range(len(u1)):
        u_vec[i] += np.tile(u1[i], (2,1))
    return u_vec # <---- moved outside the for loop

Having solved this problem, you might also be interested in knowing an alternative solution using repeat and reshape to get the desired result as

import numpy as np

n=10
u1 = np.random.rand(n)
print(u1)

u = np.repeat(u1,2).reshape((n,2,1))
print(u)

---

[0.17106854 0.7346424  0.53370937 0.39838919 0.42247593 0.61545304
 0.97014742 0.85912941 0.51137618 0.08148184]
[[[0.17106854]
  [0.17106854]]

 [[0.7346424 ]
  [0.7346424 ]]

 [[0.53370937]
  [0.53370937]]

 [[0.39838919]
  [0.39838919]]

 [[0.42247593]
  [0.42247593]]

 [[0.61545304]
  [0.61545304]]

 [[0.97014742]
  [0.97014742]]

 [[0.85912941]
  [0.85912941]]

 [[0.51137618]
  [0.51137618]]

 [[0.08148184]
  [0.08148184]]]