How to append numpy array as rows when the initial array is empty?-CodePudding

I have a function that returns an array, I want to append several of these arrays and get the mean of each row.

Basic code I want:

arr = []
for i in range(0,3):
    b = get_array()
    arr.append(b)
    
b_mean = np.mean(b, axis=0)

What I hope to achieve with it:

'''
arr = np.array([])
b0 = np.array([3, 3, 3, 5])
b1 = np.array([6, 6, 5, 7])
b3 = np.array([1, 2, 3, 4])

desired_result = np.array([[3, 3, 3, 5],
                           [6, 6, 5, 7],
                           [3, 2, 7, 3]])

desired_result_average = np.array([4, 3.66, 5, 5])
'''

I think I can achieve this by first making the array a regular list, and later converting the list back to a numpy array so I can get the mean. But this seems like a weird way...

I tried np.concatenate, np.insert, np.vstack. But these all require an initial array of the dimensions as the other arrays..

CodePudding user response：

No need to be aware of the shape of b.

arrs = []
for i in range(0,3):
    b = get_array()
    arrs.append(b)

arr = np.vstack(arrs)

b_mean = np.mean(arr, axis=0)