Mastering $group and $project

Welcome to Day 2 of Aggregation Week!

Yesterday, we learned the basics. Today, we dive deeper into Processing Data. The $group stage is where the magic happens. It allows you to pivot your data just like in Excel.

1. The Accumulators

When you group data, you usually want to calculate something.

$sum: Adds up values (or counts if you use 1).
$avg: Calculates the average.
$min / $max: Finds extremes.
$first / $last: Useful for sorted groups.
$push: Creates an array of values.

Example: E-Commerce Report

Imagine a Sales collection.

{ "item": "Laptop", "price": 1000, "category": "Electronics", "date": "2023-01-01" }
{ "item": "Mouse", "price": 50, "category": "Electronics", "date": "2023-01-02" }

Goal: Report Total Revenue and Average Price per Category.

db.sales.aggregate([
  {
    $group: {
      _id: "$category", // Group Key
      totalRevenue: { $sum: "$price" },
      averagePrice: { $avg: "$price" },
      count: { $sum: 1 } // Count how many sales
    }
  }
])

2. Dealing with Dates

Grouping by date is tricky because ISODate() includes the exact millisecond. You rarely want to group by millisecond. You want “Sales per Day” or “Sales per Month”.

Use $dateToString inside $group.

db.sales.aggregate([
  {
    $group: {
      _id: { 
        $dateToString: { format: "%Y-%m-%d", date: "$date" } 
      },
      dailySales: { $sum: "$price" }
    }
  },
  { $sort: { _id: 1 } } // Sort by date
])

3. Reshaping with `$project`

Often, the output of $group (_id, totalRevenue) looks “database-y”. The frontend wants clean JSON. Use $project to rename fields.

...
{
  $project: {
    _id: 0, // Hide the ID
    date: "$_id",
    revenue: "$dailySales"
  }
}

Output:

{ "date": "2023-01-01", "revenue": 1000 }

🧠 Daily Challenge

You have a Students collection with { name: "Alice", grade: 90, subject: "Math" }.

Calculate the Average Grade per Subject.
Find the Highest Grade in each Subject.
Use $project to make the output look like: { "subject": "Math", "average": 85, "best": 99 }.

See you on Day 3 for $lookup, where we finally join collections together! 🔗

Next Step

Next: Lookups And Joins →