How to Use reshape in NumPy: Syntax and Examples

Use numpy.reshape() to change the shape of an array without changing its data. You provide the new shape as a tuple, and NumPy returns a new array with that shape if possible.

Syntax

The reshape function changes the shape of an existing NumPy array. It takes the new shape as a tuple of integers. One dimension can be set to -1 to let NumPy calculate it automatically.

• array.reshape(new_shape): returns a new array with the specified shape.
• new_shape: tuple of integers representing the desired shape.
• -1 in new_shape: automatically calculates the dimension size.

python

numpy.reshape(a, newshape, order='C')

Example

This example shows how to reshape a 1D array of 6 elements into a 2D array with 2 rows and 3 columns.

python

import numpy as np

arr = np.array([1, 2, 3, 4, 5, 6])
reshaped_arr = arr.reshape((2, 3))
print(reshaped_arr)

Output

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

Common Pitfalls

Common mistakes when using reshape include:

• Trying to reshape to a shape that does not match the total number of elements.
• Forgetting that -1 can only be used once in the new shape.
• Modifying the original array shape without assigning or using the returned reshaped array.

python

import numpy as np

arr = np.array([1, 2, 3, 4])

# Wrong: total elements mismatch
# arr.reshape((3, 2))  # Raises ValueError

# Right: use compatible shape
reshaped = arr.reshape((2, 2))
print(reshaped)

# Using -1 to infer dimension
auto_reshaped = arr.reshape((-1, 2))
print(auto_reshaped)

Output

[[1 2]
 [3 4]]
[[1 2]
 [3 4]]

Quick Reference

Parameter	Description
a	Input array to reshape
newshape	Tuple specifying the new shape
order	'C' for row-major, 'F' for column-major (default 'C')

Key Takeaways

Use numpy.reshape() to change an array's shape without changing data.

The new shape must match the total number of elements in the array.

You can use -1 in one dimension to let NumPy calculate it automatically.

reshape() returns a new array; it does not modify the original array in place.

Always ensure the new shape is compatible with the original array size.