How to Use loc in pandas for Data Selection and Filtering

In pandas, loc is used to select rows and columns by their labels. You can specify row labels, column labels, or both inside df.loc[row_label, column_label] to get or set data based on index names.

Syntax

The basic syntax of loc is df.loc[row_indexer, column_indexer]. Here:

• row_indexer: label(s) of the row(s) you want to select.
• column_indexer: label(s) of the column(s) you want to select.

You can use single labels, lists of labels, slices, or boolean arrays for both row and column indexers.

python

df.loc[row_label, column_label]

Example

This example shows how to select specific rows and columns using loc. It demonstrates selecting a single row, multiple rows, and specific columns by their labels.

python

import pandas as pd

# Create a sample DataFrame
data = {'Name': ['Alice', 'Bob', 'Charlie', 'David'],
        'Age': [25, 30, 35, 40],
        'City': ['NY', 'LA', 'Chicago', 'Houston']}

df = pd.DataFrame(data, index=['a', 'b', 'c', 'd'])

# Select row with label 'b'
row_b = df.loc['b']

# Select rows 'a' to 'c' and columns 'Name' and 'Age'
subset = df.loc['a':'c', ['Name', 'Age']]

print("Row with label 'b':")
print(row_b)
print("\nSubset of rows 'a' to 'c' and columns 'Name' and 'Age':")
print(subset)

Output

Row with label 'b':
Name     Bob
Age       30
City      LA
Name: b, dtype: object

Subset of rows 'a' to 'c' and columns 'Name' and 'Age':
     Name  Age
a   Alice   25
b     Bob   30
c  Charlie   35

Common Pitfalls

One common mistake is confusing loc with iloc. loc uses labels, while iloc uses integer positions. Another pitfall is using labels that do not exist, which causes a KeyError. Also, slicing with loc includes the end label, unlike Python's usual slicing.

python

import pandas as pd

data = {'Value': [10, 20, 30]}

df = pd.DataFrame(data, index=['x', 'y', 'z'])

# Wrong: using integer position with loc (raises KeyError)
try:
    print(df.loc[1])
except KeyError as e:
    print(f"KeyError: {e}")

# Right: use label with loc
print(df.loc['y'])

# Slicing includes end label
print(df.loc['x':'y'])

Output

KeyError: 1
Value    20
Name: y, dtype: int64
   Value
x     10
y     20

Quick Reference

Usage	Description	Example
Select single row by label	Returns a Series for the row	df.loc['row_label']
Select multiple rows by labels	Returns DataFrame for rows	df.loc[['row1', 'row2']]
Select rows and columns	Returns subset DataFrame	df.loc['row1':'row3', ['col1', 'col2']]
Boolean indexing	Select rows where condition is True	df.loc[df['Age'] > 30]
Set values	Assign new values to subset	df.loc['a', 'Age'] = 26

Key Takeaways

Use df.loc[row_label, column_label] to select data by labels in pandas.

loc includes the end label when slicing rows or columns.

Always use labels with loc, not integer positions (use iloc for positions).

You can select single or multiple rows and columns with loc.

loc supports boolean conditions to filter rows easily.