GroupBy Aggregations

What is Aggregation?

Aggregation means combining many values into one.

Sum of all sales = aggregation
Average price = aggregation
Count of orders = aggregation

Multiple Aggregations at Once

Instead of doing sum(), mean(), count() separately, do them all together:

code.py

import pandas as pd

sales = pd.DataFrame({
    'Store': ['NYC', 'NYC', 'LA', 'LA', 'LA'],
    'Sales': [100, 150, 120, 180, 90]
})

# Get sum, mean, and count for each store
result = sales.groupby('Store')['Sales'].agg(['sum', 'mean', 'count'])
print(result)

Output:

       sum   mean  count
Store
LA     390  130.0      3
NYC    250  125.0      2

One line gives you three calculations!

Available Aggregation Functions

code.py

result = sales.groupby('Store')['Sales'].agg([
    'sum',      # Total
    'mean',     # Average
    'median',   # Middle value
    'min',      # Smallest
    'max',      # Largest
    'count',    # How many
    'std'       # Spread of values
])
print(result)

Different Aggregations for Different Columns

code.py

data = pd.DataFrame({
    'Store': ['NYC', 'NYC', 'LA', 'LA'],
    'Sales': [100, 150, 120, 180],
    'Customers': [20, 25, 30, 35]
})

# Sum for Sales, Mean for Customers
result = data.groupby('Store').agg({
    'Sales': 'sum',
    'Customers': 'mean'
})
print(result)

Output:

       Sales  Customers
Store
LA       300       32.5
NYC      250       22.5

Multiple Aggregations per Column

code.py

result = data.groupby('Store').agg({
    'Sales': ['sum', 'mean'],
    'Customers': ['min', 'max']
})
print(result)

Output:

       Sales        Customers
         sum   mean       min  max
Store
LA       300  150.0        30   35
NYC      250  125.0        20   25

Custom Aggregation Names

code.py

result = data.groupby('Store').agg(
    total_sales=('Sales', 'sum'),
    avg_sales=('Sales', 'mean'),
    total_customers=('Customers', 'sum')
)
print(result)

Output:

       total_sales  avg_sales  total_customers
Store
LA             300      150.0               65
NYC            250      125.0               45

Column names are now clear and readable.

Using Custom Functions

code.py

# Custom function: range (max - min)
def sales_range(x):
    return x.max() - x.min()

result = data.groupby('Store')['Sales'].agg(['sum', 'mean', sales_range])
print(result)

Output:

       sum   mean  sales_range
Store
LA     300  150.0           60
NYC    250  125.0           50

Lambda Functions

code.py

result = data.groupby('Store')['Sales'].agg([
    'sum',
    'mean',
    ('range', lambda x: x.max() - x.min()),
    ('above_100', lambda x: (x > 100).sum())
])
print(result)

Practice Example

code.py

import pandas as pd

# Order data
orders = pd.DataFrame({
    'Customer': ['Alice', 'Alice', 'Bob', 'Bob', 'Bob'],
    'Product': ['Laptop', 'Mouse', 'Laptop', 'Keyboard', 'Mouse'],
    'Amount': [1000, 25, 1000, 75, 25],
    'Quantity': [1, 2, 1, 1, 3]
})

print("Order Data:")
print(orders)

# Summary per customer
summary = orders.groupby('Customer').agg(
    total_spent=('Amount', 'sum'),
    avg_order=('Amount', 'mean'),
    num_orders=('Amount', 'count'),
    total_items=('Quantity', 'sum')
)
print("\nCustomer Summary:")
print(summary)

Output:

          total_spent  avg_order  num_orders  total_items
Customer
Alice           1025      512.5           2            3
Bob             1100      366.7           3            5

Reset Index After Aggregation

code.py

# Get flat DataFrame
result = orders.groupby('Customer').agg(
    total_spent=('Amount', 'sum')
).reset_index()

print(result)

Output:

  Customer  total_spent
0    Alice         1025
1      Bob         1100

Key Points

agg(['sum', 'mean']) calculates multiple stats at once
agg({'col1': 'sum', 'col2': 'mean'}) different stats per column
Use named aggregations for clear column names
Custom functions work with agg()
Use reset_index() to get regular DataFrame

Common Aggregation Functions

Function	What It Does
sum	Total of all values
mean	Average
median	Middle value
min	Smallest value
max	Largest value
count	Number of values
std	Standard deviation
var	Variance
first	First value
last	Last value

What's Next?

You learned aggregations. Next, you'll learn to group by multiple columns for detailed analysis.