How to Select Columns in NumPy Arrays Easily

To select columns in a numpy array, use slicing with array[:, column_index] where : means all rows and column_index is the column number. You can select multiple columns using a list of indices like array[:, [0, 2]].

Syntax

Use array[:, column_index] to select one column, where : means all rows and column_index is the column number. To select multiple columns, use array[:, [col1, col2, ...]] with a list of column indices.

python

array[:, column_index]
array[:, [col1, col2, ...]]

Example

This example shows how to select a single column and multiple columns from a 2D NumPy array.

python

import numpy as np

# Create a 3x4 array
arr = np.array([[10, 20, 30, 40],
                [50, 60, 70, 80],
                [90, 100, 110, 120]])

# Select the second column (index 1)
col_1 = arr[:, 1]

# Select the first and third columns (indices 0 and 2)
cols_0_2 = arr[:, [0, 2]]

print("Original array:\n", arr)
print("\nSecond column:\n", col_1)
print("\nFirst and third columns:\n", cols_0_2)

Output

Original array:
 [[ 10  20  30  40]
 [ 50  60  70  80]
 [ 90 100 110 120]]

Second column:
 [ 20  60 100]

First and third columns:
 [[ 10  30]
 [ 50  70]
 [ 90 110]]

Common Pitfalls

One common mistake is forgetting to use : for rows, which leads to errors or wrong shapes. For example, array[1] selects a row, not a column. Also, using a single index without a list returns a 1D array, while using a list returns a 2D array.

python

import numpy as np
arr = np.array([[1, 2, 3], [4, 5, 6]])

# Wrong: tries to select column without rows slice
# This returns a row, not a column
wrong = arr[1]

# Right: select second column
right = arr[:, 1]

# Wrong: single index returns 1D array
single_col = arr[:, 1]

# Right: list of indices returns 2D array
multi_cols = arr[:, [0, 2]]

print("Wrong (row):", wrong)
print("Right (column):", right)
print("Single column shape:", single_col.shape)
print("Multiple columns shape:", multi_cols.shape)

Output

Wrong (row): [4 5 6]
Right (column): [2 5]
Single column shape: (2,)
Multiple columns shape: (2, 2)

Quick Reference

Operation	Syntax	Description
Select one column	array[:, col_index]	Selects all rows for one column
Select multiple columns	array[:, [col1, col2]]	Selects all rows for listed columns
Select a row	array[row_index, :]	Selects all columns for one row
Select subarray	array[row_start:row_end, col_start:col_end]	Selects a block of rows and columns

Key Takeaways

Use array[:, column_index] to select a single column from all rows.

Use array[:, [col1, col2]] to select multiple columns at once.

Remember that array[row_index] selects a row, not a column.

Selecting a single column returns a 1D array; selecting multiple columns returns a 2D array.

Always include the row slice ':' to select columns properly.