Goal¶

This post aims to describe how to reshape an array from 1D to 2D or 2D to 1D using numpy.

Reference:

numpy.reshape

Library¶

In [1]:

import numpy as np

Create a 1D and 2D array¶

In [6]:

# 1D array
arr_1d = np.array(np.arange(0, 10))
arr_1d

Out[6]:

array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])

In [7]:

arr_1d.shape

Out[7]:

(10,)

In [12]:

# 2D array
arr_2d = np.array([np.arange(1, 20, 2), np.arange(100, 80, -2)]).T
arr_2d

Out[12]:

array([[  1, 100],
       [  3,  98],
       [  5,  96],
       [  7,  94],
       [  9,  92],
       [ 11,  90],
       [ 13,  88],
       [ 15,  86],
       [ 17,  84],
       [ 19,  82]])

In [13]:

arr_2d.shape

Out[13]:

(10, 2)

Reshape¶

reshape from 1D to 2D¶

In [9]:

# 1D with shape (10, ) to 2D with shape (2, 5)
np.reshape(arr_1d, [2, 5])

Out[9]:

array([[0, 1, 2, 3, 4],
       [5, 6, 7, 8, 9]])

In [18]:

np.reshape(arr_1d, [2, 5]).shape

Out[18]:

(2, 5)

reshape from 2D to 1D¶

In [16]:

# 2D with shape (10, 2) to 2D with shape (20, )
np.reshape(arr_2d, arr_2d.size)

Out[16]:

array([  1, 100,   3,  98,   5,  96,   7,  94,   9,  92,  11,  90,  13,
        88,  15,  86,  17,  84,  19,  82])

In [17]:

np.reshape(arr_2d, arr_2d.size).shape

Out[17]:

(20,)