How to access a NumPy array by column?

When working with datasets in Python, accessing specific columns of a NumPy array is a fundamental operation for data analysis and manipulation. NumPy provides several powerful methods to access columns efficiently.

In this article, we will explore different techniques to access columns in a NumPy array, from basic indexing to advanced boolean filtering.

Method 1: Basic Indexing

Basic indexing is the simplest way to access a column. Use the colon ":" operator to select all rows and specify the column index ?

Example

import numpy as np

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

# Access the third column (index 2)
column = array[:, 2]
print("Third column:")
print(column)

The output of the above code is ?

Third column:
[ 3  7 11]

Method 2: Fancy Indexing

Fancy indexing allows you to access multiple columns simultaneously by passing an array of column indices ?

Example

import numpy as np

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

# Access columns at indices 1 and 3
columns = array[:, [1, 3]]
print("Columns 1 and 3:")
print(columns)

The output of the above code is ?

Columns 1 and 3:
[[ 2  4]
 [ 6  8]
 [10 12]]

Method 3: Boolean Indexing

Boolean indexing allows you to select columns based on specific conditions using a boolean mask ?

Example

import numpy as np

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

# Select columns where the sum is greater than 15
condition = array.sum(axis=0) > 15
columns_filtered = array[:, condition]
print("Columns with sum > 15:")
print(columns_filtered)
print("Column sums:", array.sum(axis=0))

The output of the above code is ?

Columns with sum > 15:
[[ 2  3  4]
 [ 6  7  8]
 [10 11 12]]
Column sums: [15 18 21 24]

Method 4: Using Array Transpose

Transposing swaps rows and columns, allowing you to access columns as rows using the .T attribute ?

Example

import numpy as np

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

# Transpose and access the third column (now third row)
transposed = array.T
column = transposed[2]
print("Third column via transpose:")
print(column)

The output of the above code is ?

Third column via transpose:
[ 3  7 11]

Comparison

Conclusion

NumPy offers multiple efficient ways to access array columns. Use basic indexing for single columns, fancy indexing for multiple columns, boolean indexing for conditional selection, and transpose for column-as-row operations.