Python Pandas – Count the Observations

In Pandas, you can count the observations (rows) within groups using the groupby() method combined with count(). This is useful for analyzing the frequency of categories in your data.

Creating a Sample DataFrame

Let's start by creating a DataFrame with product information ?

import pandas as pd

# Create a DataFrame with product data
dataFrame = pd.DataFrame({
    'Product Name': ['Keyboard', 'Charger', 'SmartTV', 'Camera', 'Graphic Card', 'Earphone'],
    'Product Category': ['Computer', 'Mobile Phone', 'Electronics', 'Electronics', 'Computer', 'Mobile Phone'],
    'Quantity': [10, 50, 10, 20, 25, 50]
})

print("Original DataFrame:")
print(dataFrame)

Original DataFrame:
   Product Name Product Category  Quantity
0      Keyboard         Computer        10
1       Charger     Mobile Phone        50
2       SmartTV      Electronics        10
3        Camera      Electronics        20
4  Graphic Card         Computer        25
5      Earphone     Mobile Phone        50

Counting Observations by Group

Use groupby() to group data by category, then count() to get observation counts ?

import pandas as pd

# Create the DataFrame
dataFrame = pd.DataFrame({
    'Product Name': ['Keyboard', 'Charger', 'SmartTV', 'Camera', 'Graphic Card', 'Earphone'],
    'Product Category': ['Computer', 'Mobile Phone', 'Electronics', 'Electronics', 'Computer', 'Mobile Phone'],
    'Quantity': [10, 50, 10, 20, 25, 50]
})

# Group by Product Category and count observations
group = dataFrame.groupby("Product Category")
result = group.count()

print("Count of observations by Product Category:")
print(result)

Count of observations by Product Category:
                 Product Name  Quantity
Product Category                       
Computer                    2         2
Electronics                 2         2
Mobile Phone                2         2

Understanding the Output

The result shows that each product category has 2 observations (rows). Both Product Name and Quantity columns show the same count because count() counts non-null values in each column.

Getting Count for Specific Columns

You can also count observations for specific columns only ?

import pandas as pd

dataFrame = pd.DataFrame({
    'Product Name': ['Keyboard', 'Charger', 'SmartTV', 'Camera', 'Graphic Card', 'Earphone'],
    'Product Category': ['Computer', 'Mobile Phone', 'Electronics', 'Electronics', 'Computer', 'Mobile Phone'],
    'Quantity': [10, 50, 10, 20, 25, 50]
})

# Count only Product Name column
group = dataFrame.groupby("Product Category")
result = group['Product Name'].count()

print("Count of Product Names by Category:")
print(result)

Count of Product Names by Category:
Product Category
Computer        2
Electronics     2
Mobile Phone    2
Name: Product Name, dtype: int64

Alternative: Using size()

The size() method counts all rows in each group, including those with null values ?

import pandas as pd

dataFrame = pd.DataFrame({
    'Product Name': ['Keyboard', 'Charger', 'SmartTV', 'Camera', 'Graphic Card', 'Earphone'],
    'Product Category': ['Computer', 'Mobile Phone', 'Electronics', 'Electronics', 'Computer', 'Mobile Phone'],
    'Quantity': [10, 50, 10, 20, 25, 50]
})

# Using size() to count all rows per group
group_size = dataFrame.groupby("Product Category").size()

print("Group sizes using size():")
print(group_size)

Group sizes using size():
Product Category
Computer        2
Electronics     2
Mobile Phone    2
dtype: int64

Conclusion

Use groupby().count() to count non-null observations in each group. For counting all rows including nulls, use size() instead. This technique is essential for data analysis and understanding category distributions.