我如何在不使用循环的情况下对 3 个以上的多维数组进行元素乘法

Question

我有一个包含四个二维数组的 3 维数组

>>> print(newimagetensor) # printing the array 

[[[1.06340611e+02 1.83682746e+02 2.91655784e-02 7.70948060e+01]
  [3.74227522e+01 2.35463417e+01 4.74963539e+01 8.81179854e+01]
  [1.01175706e+02 1.37398267e+02 1.06894601e+02 1.74730973e+02]
  [5.21353237e+01 2.23919946e+02 6.98383627e+00 1.70969215e+02]]

 [[1.06412725e+02 1.42465906e+02 3.57986693e+01 5.05158797e+01]
  [2.04189865e+02 2.46906702e+02 7.99231654e+01 1.76542267e+02]
  [2.23479234e+02 2.28124699e+02 2.16862739e+01 9.95896972e+00]
  [4.33067570e+01 2.23926338e+02 2.50784426e+01 1.07382444e+02]]

 [[2.44261830e+02 1.35957148e+02 1.76428664e+02 8.04564859e+01]
  [1.75057737e+02 2.12829546e+02 4.66351072e+00 1.91286800e+02]
  [2.52159578e+02 1.90782242e+02 7.15132180e+01 2.01266229e+02]
  [2.63226317e+01 1.14212849e+02 2.31691853e+02 7.48716078e+01]]

 [[7.33827113e+01 3.31572859e+01 4.93857426e+00 1.73103061e+02]
  [5.39651696e+01 6.77143981e+01 1.25351156e+02 1.36074490e+01]
  [1.46399989e+02 3.74157866e+01 1.50272912e+02 1.78438382e+02]
  [2.60952794e+01 1.05584277e+02 1.77072040e+02 1.05615714e+02]]]

通常我可以使用循环

对这些图像进行逐元素乘法

>>> result = np.ones([4,4])   #creating a ones array equal in size to our images

>>> for i in range(len(newimagetensor)):
 result *= newimagetensor[i]                      #Multiply all the images in the newimagetensor 

>>> print(result)

输出

[[2.02834617e+08 1.17966943e+08 9.09720983e+02 5.42398960e+07]
 [7.21879586e+07 8.37855764e+07 2.21908670e+06 4.04925360e+07]
 [8.34699072e+08 2.23741421e+08 2.49119505e+07 6.24947437e+07]
 [1.55088285e+06 6.04661303e+08 7.18547308e+06 1.45176694e+08]]

但我想在不使用循环的情况下做同样的事情，而且如果可能的话，也只用两行代码。是否有相应的函数或库？

Answer 1

numpy.prod(array, axis=0) 适合我

我如何在不使用循环的情况下对 3 个以上的多维数组进行元素乘法

How can i do elementwise multiplication of more than 3 multidimensional arrays without using loop

python

arrays

numpy

matrix

elementwise-operations