Informatics Practices · Class 12 · Python Pandas · DataFrame

PandasDataFrame⏱️ 6 min read

Selecting Columns

The most everyday thing you do with a DataFrame is pull out a column. Each column comes back as a Series — the same 1-D structure you already know.

1One column at a time

Put the column name in square brackets. You can also use dot notation — but only when the name is a single word with no spaces.

one_column.py

import pandas as pd

df = pd.DataFrame({
    'Name':          ['Rinku', 'Ritu', 'Pankaj', 'Ajay', 'Aditya'],
    'English Marks': [65, 77, 72, 81, 92],
    'IP Marks':      [89, 83, 95, 99, 87],
    'Maths Marks':   [72, 69, 87, 94, 96],
}, index=['ID 1', 'ID 2', 'ID 3', 'ID 4', 'ID 5'])

# The whole DataFrame, before we pick anything out
print("--- The whole df ---")
print(df)

# Square brackets — always works
print("\n--- df['IP Marks'] ---")
print(df['IP Marks'])

# Dot notation — same column, same result
print("\n--- df.Name ---")
print(df.Name)

import pandas as pd

df = pd.DataFrame({
    'Name':          ['Rinku', 'Ritu', 'Pankaj', 'Ajay', 'Aditya'],
    'English Marks': [65, 77, 72, 81, 92],
    'IP Marks':      [89, 83, 95, 99, 87],
    'Maths Marks':   [72, 69, 87, 94, 96],
}, index=['ID 1', 'ID 2', 'ID 3', 'ID 4', 'ID 5'])

# The whole DataFrame, before we pick anything out
print("--- The whole df ---")
print(df)

# Square brackets — always works
print("\n--- df['IP Marks'] ---")
print(df['IP Marks'])

# Dot notation — same column, same result
print("\n--- df.Name ---")
print(df.Name)

Watch Out

Dot notation fails on multi-word names: df.IP Marks is a syntax error. When in doubt, use df['IP Marks'].

2Several columns at once

Pass a list of column names inside the square brackets. The columns come back in the order you list them — and the result is a DataFrame, not a Series.

many_columns.py

import pandas as pd

df = pd.DataFrame({
    'Name':          ['Rinku', 'Ritu', 'Pankaj', 'Ajay', 'Aditya'],
    'English Marks': [65, 77, 72, 81, 92],
    'IP Marks':      [89, 83, 95, 99, 87],
    'Maths Marks':   [72, 69, 87, 94, 96],
}, index=['ID 1', 'ID 2', 'ID 3', 'ID 4', 'ID 5'])

# Before: all four columns
print("--- The whole df ---")
print(df)

# After: a list of names -> a smaller DataFrame
print("\n--- df[['Name', 'Maths Marks']] ---")
print(df[['Name', 'Maths Marks']])

import pandas as pd

df = pd.DataFrame({
    'Name':          ['Rinku', 'Ritu', 'Pankaj', 'Ajay', 'Aditya'],
    'English Marks': [65, 77, 72, 81, 92],
    'IP Marks':      [89, 83, 95, 99, 87],
    'Maths Marks':   [72, 69, 87, 94, 96],
}, index=['ID 1', 'ID 2', 'ID 3', 'ID 4', 'ID 5'])

# Before: all four columns
print("--- The whole df ---")
print(df)

# After: a list of names -> a smaller DataFrame
print("\n--- df[['Name', 'Maths Marks']] ---")
print(df[['Name', 'Maths Marks']])

3Going one step further: a single value

A selected column is a Series, and a Series lets you ask for one label. So add a second pair of brackets — the row label — and you land on a single cell. Read it as column first, then row.

one_value.py

import pandas as pd

df = pd.DataFrame({
    'Name':          ['Rinku', 'Ritu', 'Pankaj', 'Ajay', 'Aditya'],
    'English Marks': [65, 77, 72, 81, 92],
    'IP Marks':      [89, 83, 95, 99, 87],
    'Maths Marks':   [72, 69, 87, 94, 96],
}, index=['ID 1', 'ID 2', 'ID 3', 'ID 4', 'ID 5'])

# Step 1: one column -> a Series
print("--- df['Name'] ---")
print(df['Name'])

# Step 2: one label out of that Series -> a single value
print("\n--- df['Name']['ID 1'] ---")
print(df['Name']['ID 1'])

# Dot notation works here too
print("\n--- df.Name['ID 1'] ---")
print(df.Name['ID 1'])

import pandas as pd

df = pd.DataFrame({
    'Name':          ['Rinku', 'Ritu', 'Pankaj', 'Ajay', 'Aditya'],
    'English Marks': [65, 77, 72, 81, 92],
    'IP Marks':      [89, 83, 95, 99, 87],
    'Maths Marks':   [72, 69, 87, 94, 96],
}, index=['ID 1', 'ID 2', 'ID 3', 'ID 4', 'ID 5'])

# Step 1: one column -> a Series
print("--- df['Name'] ---")
print(df['Name'])

# Step 2: one label out of that Series -> a single value
print("\n--- df['Name']['ID 1'] ---")
print(df['Name']['ID 1'])

# Dot notation works here too
print("\n--- df.Name['ID 1'] ---")
print(df.Name['ID 1'])

Note

This is the quick shortcut. Later, in the loc/iloc lesson, you'll meet df.at['ID 1', 'Name'] — the proper way to reach one cell, and it reads row first, then column.

Key Takeaway

df['col'] → one column as a Series. df[['a', 'b']] → several columns as a DataFrame (note the double brackets). df['col']['row'] → one value.

Quick Check

What does df[['Name', 'IP Marks']] return?

Quick Check

Which safely selects a column named 'IP Marks'?

← Previous

🔍 DataFrame Attributes

Rows & Subsets: loc / iloc 🎯