How to count frequency of itemsets in Pandas DataFrame

The value_counts() method in Pandas is used to count the frequency of unique values in a DataFrame column. This is particularly useful for analyzing categorical data and understanding data distribution patterns.

Creating the DataFrame

First, let's create a sample DataFrame with car sales data ?

import pandas as pd

# Create DataFrame
dataFrame = pd.DataFrame({
    'Car': ['BMW', 'Mercedes', 'Lamborghini', 'Audi', 'Mercedes', 'Porsche', 'Lamborghini', 'BMW'],
    'Place': ['Delhi', 'Hyderabad', 'Chandigarh', 'Bangalore', 'Hyderabad', 'Mumbai', 'Mumbai', 'Pune'],
    'UnitsSold': [95, 80, 80, 75, 92, 90, 95, 50]
})

print("DataFrame...")
print(dataFrame)

DataFrame...
          Car       Place  UnitsSold
0         BMW       Delhi         95
1    Mercedes   Hyderabad         80
2 Lamborghini  Chandigarh         80
3        Audi   Bangalore         75
4    Mercedes   Hyderabad         92
5     Porsche      Mumbai         90
6 Lamborghini      Mumbai         95
7         BMW        Pune         50

Counting Frequency Using value_counts()

Now we can count the frequency of values in each column using the value_counts() method ?

import pandas as pd

dataFrame = pd.DataFrame({
    'Car': ['BMW', 'Mercedes', 'Lamborghini', 'Audi', 'Mercedes', 'Porsche', 'Lamborghini', 'BMW'],
    'Place': ['Delhi', 'Hyderabad', 'Chandigarh', 'Bangalore', 'Hyderabad', 'Mumbai', 'Mumbai', 'Pune'],
    'UnitsSold': [95, 80, 80, 75, 92, 90, 95, 50]
})

# Counting frequency of Car column
car_counts = dataFrame['Car'].value_counts()
print("Count in column Car:")
print(car_counts)

# Counting frequency of Place column
place_counts = dataFrame['Place'].value_counts()
print("\nCount in column Place:")
print(place_counts)

# Counting frequency of UnitsSold column
units_counts = dataFrame['UnitsSold'].value_counts()
print("\nCount in column UnitsSold:")
print(units_counts)

Count in column Car:
Car
BMW            2
Lamborghini    2
Mercedes       2
Audi           1
Porsche        1
Name: count, dtype: int64

Count in column Place:
Place
Mumbai        2
Hyderabad     2
Chandigarh    1
Pune          1
Delhi         1
Bangalore     1
Name: count, dtype: int64

Count in column UnitsSold:
UnitsSold
95    2
80    2
92    1
75    1
90    1
50    1
Name: count, dtype: int64

Additional Options

The value_counts() method provides several useful parameters ?

import pandas as pd

dataFrame = pd.DataFrame({
    'Car': ['BMW', 'Mercedes', 'Lamborghini', 'Audi', 'Mercedes', 'Porsche', 'Lamborghini', 'BMW'],
    'Place': ['Delhi', 'Hyderabad', 'Chandigarh', 'Bangalore', 'Hyderabad', 'Mumbai', 'Mumbai', 'Pune']
})

# Get relative frequencies (percentages)
car_percentages = dataFrame['Car'].value_counts(normalize=True)
print("Car frequencies as percentages:")
print(car_percentages)

# Sort in ascending order
car_ascending = dataFrame['Car'].value_counts(ascending=True)
print("\nCar counts in ascending order:")
print(car_ascending)

Car frequencies as percentages:
Car
BMW            0.25
Lamborghini    0.25
Mercedes       0.25
Audi           0.125
Porsche        0.125
Name: proportion, dtype: float64

Car counts in ascending order:
Car
Audi           1
Porsche        1
BMW            2
Lamborghini    2
Mercedes       2
Name: count, dtype: int64

Key Parameters

Conclusion

The value_counts() method is essential for frequency analysis in Pandas. Use normalize=True for percentages and ascending=True to sort from lowest to highest frequency.