How to reshape (n,) array to (n,3) array? [closed]

Question

I have an array with (n,) dimension and for a project I need to have this array with (n,3) shape. This array is a set of a 3-dimensional points.

This is the code :

vertices =  np.array([line[7:] for line in content if line.startswith('vertex')]) #7 parce que 7 char dans 'vertex '
vertices = np.reshape(vertices,(vertices.size,3))

This is the error :

cannot reshape array of size 13716 into shape (13716,3)

How can I reshape the array above?

Answer 1

I am not clear about it can help you, because I don't understand very well your question, maybe I need more info. However, if each vertex have 3 coords or numbers like I write in the example, here is a solution.

# Import the dependencies
import numpy as np

# Assuming content is some like that, each vertex have three dimensions
content = [
    'vertex 1.0 2.0 3.0',
    'vertex 4.0 5.0 6.0',
    'vertex 9 10 11'
]

# Extract the coordinates of the vertices
# Split each line, take elements after 'vertex', convert to float, and create a NumPy array
vertices = np.array([line.split()[1:] for line in content if line.startswith('vertex')], dtype=float)

# Verify if the size of the array is a multiple of 3
# Reshape the array to convert it to shape (n, 3)
vertices = vertices.reshape((-1, 3))

# Print the reshaped array
print(vertices)

If this doesn't is the situation, 13746/3 = 4582. So, you can do the following:

vertices = np.reshape(vertices,(-1, 3))

NumPy calculate the necessary rows to reshape with 3 columns/dimensions.

About your question, I think that is impossible to reshape (n,) to (n,3), unless do you want to repeat the entries?