Dropping Entries

Sometimes you need to throw an item out of a Series. The drop() method removes an entry by its index label and hands you back a new Series without it.

1Drop an entry

Click the ✕ on any row to drop it and watch the Series shrink. The code updates to match what you removed.

Drop entries

Click the ✕ on any row to drop it. drop() returns a NEW Series — the labels you remove simply vanish from the result.

obj2 = obj2.drop('c')

obj2 — click ✕ to drop

a1.50

b12.75

c24.00

d35.25

e46.50

dtype: float64

Nothing dropped yet — the Series still has all 5 entries.

2Dropping a single entry

Pass the label of the entry you want gone. The result is a new Series; the original stays exactly as it was.

drop_one.py

import pandas as pd

obj2 = pd.Series([1.50, 12.75, 24.00, 35.25, 46.50],
                 index=['a', 'b', 'c', 'd', 'e'])

# Remove the entry labelled 'c'
obj2 = obj2.drop('c')
print(obj2)

import pandas as pd

obj2 = pd.Series([1.50, 12.75, 24.00, 35.25, 46.50],
                 index=['a', 'b', 'c', 'd', 'e'])

# Remove the entry labelled 'c'
obj2 = obj2.drop('c')
print(obj2)

3Dropping several at once

To remove more than one entry, pass a list of labels.

drop_many.py

import pandas as pd

s = pd.Series([10, 20, 30, 40, 50], index=['a', 'b', 'c', 'd', 'e'])

# Drop two labels in one go
print(s.drop(['b', 'd']))

import pandas as pd

s = pd.Series([10, 20, 30, 40, 50], index=['a', 'b', 'c', 'd', 'e'])

# Drop two labels in one go
print(s.drop(['b', 'd']))

Note

By default drop() returns a new Series and leaves the original alone. To change the original itself, pass inplace=True: s.drop('c', inplace=True).

Watch Out

The label must actually exist — dropping a label that isn't in the Series raises a KeyError.

Quick Check

s has labels a, b, c, d. What does s.drop(['a', 'c']) return?

Quick Check

How do you make drop() modify the original Series instead of returning a copy?

← Previous

🔀 Reindexing

Series Attributes 🔍