Reshaping Arrays in NumPy

Reshaping Arrays

In NumPy, the reshape() function allows you to change the shape of an array without changing its data. This is especially useful when preparing data for matrix operations, machine learning models, or image processing tasks.

For example, you can take a 1D array and reshape it into a 2D array with rows and columns.

import numpy as np

# Create a 1D array with 6 elements
arr = np.array([1, 2, 3, 4, 5, 6])

# Reshape into a 2D array (2 rows, 3 columns)
reshaped = arr.reshape(2, 3)

print(reshaped)

import numpy as np

# Create a 1D array with 6 elements
arr = np.array([1, 2, 3, 4, 5, 6])

# Reshape into a 2D array (2 rows, 3 columns)
reshaped = arr.reshape(2, 3)

print(reshaped)

How It Works:

• arr is a 1D array with 6 elements.
• reshape(2, 3) creates a new view of the array with 2 rows and 3 columns.
• No data is lost — just reorganized.
• The product of the dimensions in the new shape (2 × 3 = 6) must match the original number of elements in the array.

Output

[[1 2 3]
 [4 5 6]]

[[1 2 3]
 [4 5 6]]

You can reshape arrays into any shape, as long as the total number of elements stays the same. For example, the array above (6 elements) could also be reshaped into a 3x2 or 6x1 array.

print(arr.reshape(3, 2))  # 3 rows, 2 columns
print(arr.reshape(6, 1))  # 6 rows, 1 column

print(arr.reshape(3, 2))  # 3 rows, 2 columns
print(arr.reshape(6, 1))  # 6 rows, 1 column

If the reshape isn’t possible (for example, if the new shape doesn’t match the number of elements), NumPy will raise an error.

💡 Tip: You can use -1 to let NumPy automatically calculate one dimension for you:

arr.reshape(1, -1)  # NumPy will infer the correct number of columns

arr.reshape(1, -1)  # NumPy will infer the correct number of columns

This is handy when you know how many rows (or columns) you want, but not the other dimension.

Example 1: Reshape a 3D Array to 2D

In this example, we'll take a 3D array and reshape it into a 2D array. This is useful when flattening grouped data into a tabular format for easier processing or analysis. We'll use the .reshape() method to achieve this.

import numpy as np

# Create a 3D array with shape (2, 3, 2)
arr = np.array([
  [[1, 2], [3, 4], [5, 6]],
  [[7, 8], [9, 10], [11, 12]]
])

# Reshape it into a 2D array
reshaped = arr.reshape(6, 2)

print("Original shape:", arr.shape)
print("Reshaped shape:", reshaped.shape)
print("Reshaped array:")
print(reshaped)

import numpy as np

# Create a 3D array with shape (2, 3, 2)
arr = np.array([
  [[1, 2], [3, 4], [5, 6]],
  [[7, 8], [9, 10], [11, 12]]
])

# Reshape it into a 2D array
reshaped = arr.reshape(6, 2)

print("Original shape:", arr.shape)
print("Reshaped shape:", reshaped.shape)
print("Reshaped array:")
print(reshaped)

How It Works:

• arr is a 3D array with shape (2, 3, 2), meaning 2 blocks of 3 rows and 2 columns.
• arr.reshape(6, 2) flattens it into a 2D array with 6 rows and 2 columns.
• The total number of elements remains constant: 2 × 3 × 2 = 12.

Output

Original shape: (2, 3, 2)
Reshaped shape: (6, 2)
Reshaped array:
[[ 1  2]
 [ 3  4]
 [ 5  6]
 [ 7  8]
 [ 9 10]
 [11 12]]

Original shape: (2, 3, 2)
Reshaped shape: (6, 2)
Reshaped array:
[[ 1  2]
 [ 3  4]
 [ 5  6]
 [ 7  8]
 [ 9 10]
 [11 12]]

💡 Tip: When reshaping from higher to lower dimensions, think of it as “unstacking” the nested structure into a flat, row-wise layout.

Frequently Asked Questions