CODEWITHSUNDEEP
pythonarrays
Sam
Sam

Reputation: 13

Appending np.arrays to a blank array in Python

I am trying to save the results from a loop in a np.array.

    import numpy as np
    p=np.array([])
    points= np.array([[3,0,0],[-1,0,0]])

    for i in points:
       for j in points:
          if j[0]!=0:
             n=i+j
       p= np.append(p,n)

However the resulting array is a 1D array of 6 members.

    [2.  0.  0. -2.  0.  0.]

Instead I am looking for, but have been unable to produce:

    [[2,0,0],[-2,0,0]]

Is there any way to get the result above?

Thank you.

Upvotes: 1

Views: 63

Answers (2)

Lanting
Lanting

Reputation: 3068

What you're looking for is vertically stacking your results:

import numpy as np
p=np.empty((0,3))
points= np.array([[3,0,0],[-1,0,0]])

for i in points:
  for j in points:
    if j[0]!=0:
      n=i+j
  p= np.vstack((p,n))
print p

which gives:

[[ 2.  0.  0.]
 [-2.  0.  0.]]

Although you could also reshape your result afterwards:

import numpy as np
p=np.array([])
points= np.array([[3,0,0],[-1,0,0]])

for i in points:
  for j in points:
    if j[0]!=0:
      n=i+j
  p= np.append(p,n)
p=np.reshape(p,(-1,3))
print p

Which gives the same result .

I must warn you, hovever, that your code fails if j[0]!=0 as that would make n undefined...

np.vstack np.empty np.reshape

Upvotes: 0

NPE
NPE

Reputation: 500177

One possibility is to turn p into a list, and convert it into a NumPy array right at the end:

p = []
for i in points:
   ...
   p.append(n)
p = np.array(p)

Upvotes: 2

Related Questions