How to Find Bottom N Values in pandas DataFrame or Series

To find the bottom n values in a pandas DataFrame or Series, use the nsmallest(n, columns) method for DataFrames or nsmallest(n) for Series. Alternatively, you can sort the data with sort_values() and then select the first n rows.

Syntax

For DataFrame:

• df.nsmallest(n, columns): Returns the rows with the smallest n values in the specified columns.
• df.sort_values(by=columns).head(n): Sorts the DataFrame by columns ascending and selects the first n rows.

For Series:

• series.nsmallest(n): Returns the smallest n values in the Series.
• series.sort_values().head(n): Sorts the Series ascending and selects the first n values.

python

df.nsmallest(n, columns)
series.nsmallest(n)
df.sort_values(by=columns).head(n)
series.sort_values().head(n)

Example

This example shows how to find the bottom 3 values in a DataFrame column and in a Series using nsmallest() and sort_values().

python

import pandas as pd

# Create a sample DataFrame
data = {'Name': ['Alice', 'Bob', 'Charlie', 'David', 'Eva'],
        'Score': [88, 92, 79, 93, 85]}
df = pd.DataFrame(data)

# Find bottom 3 scores using nsmallest
bottom_3_nsmallest = df.nsmallest(3, 'Score')

# Find bottom 3 scores using sort_values
bottom_3_sorted = df.sort_values(by='Score').head(3)

# Create a Series
scores = pd.Series([88, 92, 79, 93, 85], index=['Alice', 'Bob', 'Charlie', 'David', 'Eva'])

# Bottom 3 values in Series
bottom_3_series = scores.nsmallest(3)

print("Bottom 3 rows in DataFrame using nsmallest:")
print(bottom_3_nsmallest)
print("\nBottom 3 rows in DataFrame using sort_values:")
print(bottom_3_sorted)
print("\nBottom 3 values in Series:")
print(bottom_3_series)

Output

Bottom 3 rows in DataFrame using nsmallest:
      Name  Score
2  Charlie     79
4      Eva     85
0    Alice     88

Bottom 3 rows in DataFrame using sort_values:
      Name  Score
2  Charlie     79
4      Eva     85
0    Alice     88

Bottom 3 values in Series:
Charlie    79
Eva        85
Alice      88
dtype: int64

Common Pitfalls

1. Forgetting to specify the column in nsmallest() for DataFrames: You must provide the column name to find bottom values in that column.

2. Using nsmallest() on DataFrame without columns argument causes error.

3. Confusing nsmallest() with nlargest() which finds top values.

python

import pandas as pd

df = pd.DataFrame({'A': [5, 2, 9], 'B': [1, 3, 7]})

# Wrong: missing column argument
# df.nsmallest(2)  # This will raise TypeError

# Right: specify column
bottom_2 = df.nsmallest(2, 'A')
print(bottom_2)

Output

   A  B
1  2  3
0  5  1

Quick Reference

• df.nsmallest(n, 'col'): Fast way to get bottom n rows by column.
• df.sort_values('col').head(n): Sort then slice to get bottom n.
• series.nsmallest(n): Bottom n values in Series.
• series.sort_values().head(n): Sort Series ascending and take first n.

Key Takeaways

Use df.nsmallest(n, column) to quickly find bottom n rows by a column in a DataFrame.

For Series, use series.nsmallest(n) to get bottom n values easily.

You can also sort with sort_values() and then select the first n rows or values.

Always specify the column name when using nsmallest() on DataFrames.

nsmallest() is more efficient than sorting when you only need a few bottom values.