形状不匹配Numpy

Question

I am continously getting the error:

"(shapes (10, 1), (10,) mismatch)"

when doing a NumPy operation and I am somewhat confused.

Wouldn't (10,1) and (10,) be identical shapes? And if for whatever reason this is not valid, is there a way to convert (10,1) to (10,)? I cannot seem to find it in the NumPy doucmentation.

Thanks

Answer 1

The difference between (10,1) and (10,) is the dimensions of your vector. The first ( (10,1) ) is a two-dimensional array while the latter ( (10,) ) is one-dimensional:

>>> import numpy as np
>>> x = np.zeros((10,1))
>>> x
array([[ 0.],
       [ 0.],
       [ 0.],
       [ 0.],
       [ 0.],
       [ 0.],
       [ 0.],
       [ 0.],
       [ 0.],
       [ 0.]])
>>> x.shape = (10,)
>>> x
array([ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.])

So long as the total number of elements remain the same, you can re-shape your array however you'd like via numpy.ndarray.shape

Answer 2

In numpy , (10, 1), (10,) are not the same at all:

• (10, 1) is a two dimensional array, with a single column.

• (10, ) is a one dimensional array

If you have an array a , and print out len(a.shape), you'll see the difference.