如何更优雅地构造2的幂的numpy数组？

Question

I would like to figure out how i can create this array without putting every single value in per hand.

Is there a way how i can use the information that every value is the doubled value of its predecessor, except for the first?

My Code is as follows:

import numpy as np

Matrix = np.array([1,2,4,8,16,32,64,128,256]).reshape (3,3)

print(Matrix)

Answer 1

You can use np.arange , and take advantage of the fact that they are powers of 2 :

2**np.arange(9).reshape(-1, 3)

array([[  1,   2,   4],
       [  8,  16,  32],
       [ 64, 128, 256]], dtype=int32)

Answer 2

You could also do something like this:

var myRandomArray = [1];
var i = 1;
var num = 1;
while (i < 9) {
  myRandomArray.push(num = num * 2);
  i = i + 1;
}

This is written in JavaScript. For Python, just switch what you need around, the main idea is still there. I believe in Python, it is append instead of push.

Answer 3

You could use np.vander :

np.vander([2], 9, True).reshape(3, 3)
# array([[  1,   2,   4],
#        [  8,  16,  32],
#        [ 64, 128, 256]])

Answer 4

Here's a jury-rigged solution:

In [10]: total_num = 9 

In [11]: np.array([2**n for n in range(0, total_num)]).reshape(3, -1) 
Out[11]: 
array([[  1,   2,   4],
       [  8,  16,  32],
       [ 64, 128, 256]])