Python – Sort grouped Pandas dataframe by group size?

To group Pandas data frame, we use groupby(). To sort grouped data frames in ascending or descending order, use sort_values(). The size() method is used to get the data frame size.

Steps Involved

The steps included in sorting the pandas data frame by its group size are as follows ?

• Importing the pandas library and creating a Pandas DataFrame.

• Grouping the columns by using the groupby() function and sorting the values by using sort_values() in descending order.

• Sorting the values in ascending order by using the sort_values() function.

Creating a Pandas DataFrame

First import the Pandas library, and create a Pandas data frame ?

import pandas as pd

# DataFrame with one of the columns as Reg_Price
dataFrame = pd.DataFrame({
    "Car": ['BMW', 'Lexus', 'Audi', 'Mercedes', 'Jaguar', 'Bentley'],
    "Reg_Price": [1000, 1400, 1000, 900, 1700, 900],
})

print("DataFrame...\n", dataFrame)

DataFrame...
       Car  Reg_Price
0      BMW       1000
1    Lexus       1400
2     Audi       1000
3 Mercedes        900
4   Jaguar       1700
5  Bentley        900

Grouping and Sorting by Group Size

Sorting in Descending Order

To group according to the Reg_Price column and sort in descending order by setting ascending=False ?

import pandas as pd

dataFrame = pd.DataFrame({
    "Car": ['BMW', 'Lexus', 'Audi', 'Mercedes', 'Jaguar', 'Bentley'],
    "Reg_Price": [1000, 1400, 1000, 900, 1700, 900],
})

# Group by Reg_Price and sort group sizes in descending order
result_desc = dataFrame.groupby('Reg_Price').size().sort_values(ascending=False)
print("Sorted in Descending order...\n", result_desc)

Sorted in Descending order...
Reg_Price
1000    2
900     2
1700    1
1400    1
dtype: int64

Sorting in Ascending Order

Group according to the Reg_Price column and sort in ascending order by setting ascending=True ?

import pandas as pd

dataFrame = pd.DataFrame({
    "Car": ['BMW', 'Lexus', 'Audi', 'Mercedes', 'Jaguar', 'Bentley'],
    "Reg_Price": [1000, 1400, 1000, 900, 1700, 900],
})

# Group by Reg_Price and sort group sizes in ascending order
result_asc = dataFrame.groupby('Reg_Price').size().sort_values(ascending=True)
print("Sorted in Ascending order...\n", result_asc)

Sorted in Ascending order...
Reg_Price
1400    1
1700    1
900     2
1000    2
dtype: int64

Complete Example

In this comprehensive example, the DataFrame contains columns Car and Reg_Price. We group by the Reg_Price column using the groupby() function, calculate the size of each group, then sort these group sizes in both descending and ascending order using the sort_values() function ?

import pandas as pd

# DataFrame with one of the columns as Reg_Price
dataFrame = pd.DataFrame({
    "Car": ['BMW', 'Lexus', 'Audi', 'Mercedes', 'Jaguar', 'Bentley'],
    "Reg_Price": [1000, 1400, 1000, 900, 1700, 900]
})

print("DataFrame...\n", dataFrame)

# Group according to Reg_Price column and sort in descending order
print("\nSorted in Descending order...")
print(dataFrame.groupby('Reg_Price').size().sort_values(ascending=False))

# Group according to Reg_Price column and sort in ascending order
print("\nSorted in Ascending order...")
print(dataFrame.groupby('Reg_Price').size().sort_values(ascending=True))

DataFrame...
       Car  Reg_Price
0      BMW       1000
1    Lexus       1400
2     Audi       1000
3 Mercedes        900
4   Jaguar       1700
5  Bentley        900

Sorted in Descending order...
Reg_Price
1000    2
900     2
1700    1
1400    1
dtype: int64

Sorted in Ascending order...
Reg_Price
1400    1
1700    1
900     2
1000    2
dtype: int64

Comparison

Conclusion

Use groupby().size().sort_values() to sort groups by their size. Set ascending=False for largest groups first, or ascending=True for smallest groups first. This technique is useful for identifying the most or least common categories in your data.