Iterating over a DataFrameOptional Reading

Sometimes you need to visit every row (or column) one at a time. Rather than write a statement for each, you iterate — a for loop that hands you one piece per turn.

1Row by row with iterrows()

df.iterrows() yields a (index, Series) pair each turn: the row's label and all its values as a Series. Perfect for "do something with each student".

iterrows.py

import pandas as pd

df = pd.DataFrame({
    'Name':    ['Rinku', 'Ritu', 'Pankaj'],
    'English': [65, 77, 72],
    'IP':      [89, 83, 95],
}, index=['ID 1', 'ID 2', 'ID 3'])

for row_index, row in df.iterrows():
    print("Row index:", row_index)
    print(row)
    print("-" * 20)

import pandas as pd

df = pd.DataFrame({
    'Name':    ['Rinku', 'Ritu', 'Pankaj'],
    'English': [65, 77, 72],
    'IP':      [89, 83, 95],
}, index=['ID 1', 'ID 2', 'ID 3'])

for row_index, row in df.iterrows():
    print("Row index:", row_index)
    print(row)
    print("-" * 20)

Tip

Inside the loop, grab one value with row['IP'] — because row is just a Series. Try adding print(row_index, row['Name'], row['IP']).

2Column by column with items()

df.items() goes the other way — one (column name, Series) pair per turn, walking across the columns.

items.py

import pandas as pd

df = pd.DataFrame({
    'Name':    ['Rinku', 'Ritu', 'Pankaj'],
    'English': [65, 77, 72],
    'IP':      [89, 83, 95],
}, index=['ID 1', 'ID 2', 'ID 3'])

for col_name, col in df.items():
    print("Column:", col_name)
    print(col)
    print("-" * 20)

import pandas as pd

df = pd.DataFrame({
    'Name':    ['Rinku', 'Ritu', 'Pankaj'],
    'English': [65, 77, 72],
    'IP':      [89, 83, 95],
}, index=['ID 1', 'ID 2', 'ID 3'])

for col_name, col in df.items():
    print("Column:", col_name)
    print(col)
    print("-" * 20)

Note

Older material (and the RM) calls the column version iteritems(). Modern Pandas renamed it to items() — they do the same thing.

Key Takeaway

iterrows() → one row (index, Series) per turn. items() → one column (name, Series) per turn. Either way, each piece is a Series you already know how to use.

Quick Check

Each turn of `for i, r in df.iterrows()`, what is r?

Quick Check

Which loop visits the DataFrame column by column?

← Previous

☑️ Boolean Indexing

head/tail & Statistics 📈