การรวมค่าเมตริก

What if you could ask "how much total sales did we have last month?" or "what's the average price of our products?" Aggregations in Elasticsearch let you do exactly that!

They provide powerful analytical capabilities beyond simple search. Think of them as SQL's GROUP BY and aggregate functions, but for your search data.

Metric aggregations are the simplest type of aggregation. They calculate a single metric (like a sum, average, min, or max) over a set of documents.

They give you numerical summaries of your data, helping you quickly understand key statistics and trends without retrieving all individual documents.

To demonstrate metric aggregations, let's index some sample sales data. We'll use documents with a price field.

Run this curl command to add our first product:

Great! To have more data for our aggregations, let's add two more products. Imagine you've run similar curl commands for these:

• sales_data/_doc/2: {"product": "Mouse", "price": 25, "quantity": 2}
• sales_data/_doc/3: {"product": "Keyboard", "price": 75, "quantity": 1}

Now we have a small dataset to work with!

The avg aggregation calculates the arithmetic mean of a numeric field. It's perfect for finding the average price of products, average age, or average score.

Let's find the average price of our indexed products:

The sum aggregation calculates the total sum of a numeric field's values. This is ideal for finding total sales, total quantity, or total revenue.

Let's calculate the total price of all indexed products:

The min and max aggregations return the smallest and largest values of a numeric field, respectively. They help identify outliers or boundary values in your data.

Let's find the cheapest and most expensive product prices:

The value_count aggregation counts the number of documents that have a value for a specific field. It's different from a document count because it only considers documents where the field exists and is not null.

Let's count how many products have a 'price' field:

The stats aggregation is a convenience aggregation that computes min, max, sum, avg, and count (of values) all at once for a numeric field.

It saves you from writing five separate aggregations and is great for a quick overview!

Consider a scenario where you have a list of product prices: 100, 50, 200, 50, null.

What would be the result of a value_count aggregation on the 'price' field?

In this lesson, you've learned about Elasticsearch metric aggregations, powerful tools for summarizing your data:

• avg: Calculates the arithmetic mean.
• sum: Computes the total sum.
• min & max: Finds the smallest and largest values.
• value_count: Counts non-null field values.
• stats: A convenient way to get all basic stats at once.

These aggregations are fundamental for understanding the numerical characteristics of your data and are often combined for more complex analysis.