Docs Menu
Docs Home
/
Languages
/
Python
/
PyMongo Driver
/
Aggregation

Group and Total

On this page

  • Introduction
  • Aggregation Task Summary
  • Before You Get Started
  • Tutorial
  • Add a match stage for orders in 2020
  • Add a sort stage to sort by order date
  • Add a group stage to group by email address
  • Add a sort stage to sort by first order date
  • Add a set stage to display the email address
  • Add an unset stage to remove unneeded fields
  • Run the aggregation pipeline
  • Interpret results

Introduction

In this tutorial, you can learn how to use PyMongo to construct an aggregation pipeline, perform the aggregation on a collection, and print the results by completing and running a sample app. This aggregation performs the following operations:

  • Matches a subset of documents by a field value

  • Groups documents by common field values

  • Adds computed fields to each result document

Aggregation Task Summary

This tutorial demonstrates how to group and analyze customer order data. The results show the list of customers who purchased items in 2020 and includes each customer's order history for 2020.

This example uses one collection, orders, which contains documents describing individual product orders. Since each order can correspond to only one customer, the order documents are grouped by the customer_id field, which contains customer email addresses.

Before You Get Started

Before you begin following an aggregation tutorial, you must set up a new Python app. You can use this app to connect to a MongoDB deployment, insert sample data into MongoDB, and run the aggregation pipeline in each tutorial.

Tip

To learn how to install the driver and connect to MongoDB, see Get Started with PyMongo

Once you install the driver, create a file called agg_tutorial.py. Paste the following code in this file to create an app template for the aggregation tutorials:

from pymongo import MongoClient
# Replace the placeholder with your connection string.
uri = "<connection string>"
client = MongoClient(uri)
try:
    agg_db = client["agg_tutorials_db"]
    # Get a reference to relevant collections.
    # ... some_coll =
    # ... another_coll =
    # Delete any existing documents in collections.
    # ... some_coll.delete_many({})
    # Insert sample data into the collection or collections.
    # ... some_data = [...]
    # ... some_coll.insert_many(some_data)
    # Create an empty pipeline array.
    pipeline = []
    # Add code to create pipeline stages.
    # ... pipeline.append({...})
    # Run the aggregation.
    # ... aggregation_result = ...
    # Print the aggregation results.
    for document in aggregation_result:
        print(document)
finally:
    client.close()

Important

In the preceding code, read the code comments to find the sections of the code that you must modify for the tutorial you are following.

If you attempt to run the code without making any changes, you will encounter a connection error.

For every tutorial, you must replace the connection string placeholder with your deployment's connection string. To learn how to locate your deployment's connection string, see Create a Connection String.

For example, if your connection string is "mongodb+srv://mongodb-example:27017", your connection string assignment resembles the following:

uri = "mongodb+srv://mongodb-example:27017";

To run the completed file after you modify the template for a tutorial, run the following command in your shell:

python3 agg_tutorial.py

After you set up the app, access the orders collection by adding the following code to the application:

orders_coll = agg_db["orders"]

Delete any existing data and insert sample data into the orders collection as shown in the following code:

orders_coll.delete_many({})
order_data = [
    {
        "customer_id": "elise_smith@myemail.com",
        "orderdate": datetime(2020, 5, 30, 8, 35, 52),
        "value": 231
    },
    {
        "customer_id": "elise_smith@myemail.com",
        "orderdate": datetime(2020, 1, 13, 9, 32, 7),
        "value": 99
    },
    {
        "customer_id": "oranieri@warmmail.com",
        "orderdate": datetime(2020, 1, 1, 8, 25, 37),
        "value": 63
    },
    {
        "customer_id": "tj@wheresmyemail.com",
        "orderdate": datetime(2019, 5, 28, 19, 13, 32),
        "value": 2
    },
    {
        "customer_id": "tj@wheresmyemail.com",
        "orderdate": datetime(2020, 11, 23, 22, 56, 53),
        "value": 187
    },
    {
        "customer_id": "tj@wheresmyemail.com",
        "orderdate": datetime(2020, 8, 18, 23, 4, 48),
        "value": 4
    },
    {
        "customer_id": "elise_smith@myemail.com",
        "orderdate": datetime(2020, 12, 26, 8, 55, 46),
        "value": 4
    },
    {
        "customer_id": "tj@wheresmyemail.com",
        "orderdate": datetime(2021, 2, 28, 7, 49, 32),
        "value": 1024
    },
    {
        "customer_id": "elise_smith@myemail.com",
        "orderdate": datetime(2020, 10, 3, 13, 49, 44),
        "value": 102
    }
]
orders_coll.insert_many(order_data)

Tutorial

1

Add a match stage for orders in 2020

First, add a $match stage that matches orders placed in 2020:

pipeline.append({
    "$match": {
        "orderdate": {
            "$gte": datetime(2020, 1, 1, 0, 0, 0),
            "$lt": datetime(2021, 1, 1, 0, 0, 0)
        }
    }
})
2

Add a sort stage to sort by order date

Next, add a $sort stage to set an ascending sort on the orderdate field to surface the earliest 2020 purchase for each customer in the next stage:

pipeline.append({
    "$sort": {
        "orderdate": 1
    }
})
3

Add a group stage to group by email address

Add a $group stage to group orders by the value of the customer_id field. In this stage, add aggregation operations that create the following fields in the result documents:

  • first_purchase_date: the date of the customer's first purchase

  • total_value: the total value of all the customer's purchases

  • total_orders: the total number of the customer's purchases

  • orders: the list of all the customer's purchases, including the date and value of each purchase

pipeline.append({
    "$group": {
        "_id": "$customer_id",
        "first_purchase_date": {"$first": "$orderdate"},
        "total_value": {"$sum": "$value"},
        "total_orders": {"$sum": 1},
        "orders": {"$push": {"orderdate": "$orderdate", "value": "$value"}}
    }
})
4

Add a sort stage to sort by first order date

Next, add another $sort stage to set an ascending sort on the first_purchase_date field:

pipeline.append({
    "$sort": {
        "first_purchase_date": 1
    }
})
5

Add a set stage to display the email address

Add a $set stage to recreate the customer_id field from the values in the _id field that were set during the $group stage:

pipeline.append({
    "$set": {
        "customer_id": "$_id"
    }
})
6

Add an unset stage to remove unneeded fields

Finally, add an $unset stage. The $unset stage removes the _id field from the result documents:

pipeline.append({"$unset": ["_id"]})
7

Run the aggregation pipeline

Add the following code to the end of your application to perform the aggregation on the orders collection:

aggregation_result = orders_coll.aggregate(pipeline)

Finally, run the following command in your shell to start your application:

python3 agg_tutorial.py
8

Interpret results

The aggregation returns the following summary of customers' orders from 2020:

{
  'first_purchase_date': datetime.datetime(2020, 1, 1, 8, 25, 37),
  'total_value': 63,
  'total_orders': 1,
  'orders': [ { 'orderdate': datetime.datetime(2020, 1, 1, 8, 25, 37), 'value': 63 } ],
  'customer_id': 'oranieri@warmmail.com'
}
{
  'first_purchase_date': datetime.datetime(2020, 1, 13, 9, 32, 7),
  'total_value': 436,
  'total_orders': 4,
  'orders': [
    { 'orderdate': datetime.datetime(2020, 1, 13, 9, 32, 7), 'value': 99 },
    { 'orderdate': datetime.datetime(2020, 5, 30, 8, 35, 52), 'value': 231 },
    { 'orderdate': datetime.datetime(2020, 10, 3, 13, 49, 44), 'value': 102 },
    { 'orderdate': datetime.datetime(2020, 12, 26, 8, 55, 46), 'value': 4 }
  ],
  'customer_id': 'elise_smith@myemail.com'
}
{
  'first_purchase_date': datetime.datetime(2020, 8, 18, 23, 4, 48),
  'total_value': 191,
  'total_orders': 2,
  'orders': [
    { 'orderdate': datetime.datetime(2020, 8, 18, 23, 4, 48), 'value': 4 },
    { 'orderdate': datetime.datetime(2020, 11, 23, 22, 56, 53), 'value': 187 }
  ],
  'customer_id': 'tj@wheresmyemail.com'
}

The result documents contain details from all the orders from a given customer, grouped by the customer's email address.

To view the complete code for this tutorial, see the Completed Group and Total App on GitHub.

Back

Filtered Subset

Next

Unpack Arrays and Group