How to Transpose Array in NumPy: Simple Guide

To transpose an array in NumPy, use the .T attribute or the np.transpose() function. Both methods flip the array's axes, turning rows into columns and vice versa.

Syntax

There are two common ways to transpose arrays in NumPy:

• array.T: A simple attribute that returns the transposed array.
• np.transpose(array, axes=None): A function that returns a view of the array with axes permuted. If axes is not given, it reverses the axes.

python

import numpy as np

# Using .T attribute
transposed_array = array.T

# Using np.transpose function
transposed_array_func = np.transpose(array, axes=None)

Example

This example shows how to transpose a 2D NumPy array using both .T and np.transpose(). The output demonstrates that rows become columns.

python

import numpy as np

array = np.array([[1, 2, 3], [4, 5, 6]])

# Transpose using .T
transposed = array.T

# Transpose using np.transpose()
transposed_func = np.transpose(array)

print("Original array:\n", array)
print("\nTransposed with .T:\n", transposed)
print("\nTransposed with np.transpose():\n", transposed_func)

Output

Original array:
 [[1 2 3]
 [4 5 6]]

Transposed with .T:
 [[1 4]
 [2 5]
 [3 6]]

Transposed with np.transpose():
 [[1 4]
 [2 5]
 [3 6]]

Common Pitfalls

Common mistakes when transposing arrays include:

• Trying to transpose a 1D array, which has no effect because it has only one axis.
• Confusing .T with copying; transposing returns a view, not a copy.
• Using np.transpose() with incorrect axes order, which can lead to unexpected shapes.

python

import numpy as np

# 1D array transpose has no effect
arr_1d = np.array([1, 2, 3])
print("1D array transpose with .T:", arr_1d.T)

# Using np.transpose with wrong axes
arr_3d = np.arange(8).reshape(2,2,2)
# Correct transpose (reverse axes)
correct = np.transpose(arr_3d)
# Wrong axes order (may cause confusion)
wrong = np.transpose(arr_3d, axes=(1,0,2))

print("\nOriginal shape:", arr_3d.shape)
print("Correct transpose shape:", correct.shape)
print("Wrong transpose shape:", wrong.shape)

Output

1D array transpose with .T: [1 2 3]

Original shape: (2, 2, 2)
Correct transpose shape: (2, 2, 2)
Wrong transpose shape: (2, 2, 2)

Quick Reference

Method	Description	Example
.T	Simple attribute to transpose array axes	array.T
np.transpose()	Function to permute axes, default reverses axes	np.transpose(array)
np.transpose(array, axes=...)	Transpose with custom axes order	np.transpose(array, axes=(1,0))

Key Takeaways

Use .T for quick transposing of 2D arrays in NumPy.

Use np.transpose() to transpose arrays with more than two dimensions or to specify axes order.

Transposing a 1D array does not change it because it has only one axis.

Both .T and np.transpose() return views, not copies, so changes affect the original array.

Be careful with the axes parameter in np.transpose() to avoid unexpected shapes.