Why Cardinality aggregation in Elasticsearch? - Purpose & Use Cases

The Big Idea

What if you could count millions of unique visitors in seconds without breaking a sweat?

The Scenario

Imagine you have a huge list of customer visits to a store, and you want to count how many unique customers came in. Doing this by checking each record one by one and remembering every customer manually is like trying to count every unique face in a busy crowd without any help.

The Problem

Manually tracking unique items in large data is slow and uses a lot of memory. It's easy to make mistakes, miss some entries, or double count. When data grows, this method becomes impossible to manage efficiently.

The Solution

Cardinality aggregation in Elasticsearch quickly counts unique values using smart algorithms. It handles huge data sets efficiently without needing to store every single item, saving time and memory.

Before vs After

✗ Before

Loop through all records, add each customer ID to a list if not already there, then count the list length.

✓ After

{
  "aggs": {
    "unique_customers": {
      "cardinality": {
        "field": "customer_id.keyword"
      }
    }
  }
}

What It Enables

It lets you quickly find the number of unique items in massive data sets without slowing down your system.

Real Life Example

A store owner wants to know how many different customers visited last month to understand their reach, without manually checking millions of sales records.

Key Takeaways

Manual counting of unique items is slow and error-prone.

Cardinality aggregation efficiently estimates unique counts in big data.

This method saves time and memory, making analysis faster and easier.

Practice

1.

What does the cardinality aggregation in Elasticsearch do?

easy

A. Calculates the sum of values in a field

B. Sorts documents by a field

C. Finds the average value of a field

D. Counts the number of unique values in a field

Solution

1. Step 1: Understand the purpose of cardinality aggregation

   Cardinality aggregation is designed to count unique values in a specified field.

2. Step 2: Compare with other aggregations

   Sum and average aggregations calculate totals and averages, not unique counts. Sorting is unrelated.

3. Final Answer:

   Counts the number of unique values in a field -> Option D

4. Quick Check:

   Cardinality = Unique count [OK]

Hint: Cardinality = count unique values fast [OK]

Common Mistakes:

• Confusing cardinality with sum or average
• Thinking it returns exact counts always
• Mixing it with sorting or filtering

2.

Which of the following is the correct syntax to define a cardinality aggregation on the field user_id in Elasticsearch?

{
  "aggs": {
    "unique_users": {
      "cardinality": {
        "field": "user_id"
      }
    }
  }
}

easy

A. { "aggs": { "unique_users": { "cardinality": { "field": "user_id" } } } }

B. { "aggs": { "unique_users": { "terms": { "field": "user_id" } } } }

C. { "aggs": { "unique_users": { "sum": { "field": "user_id" } } } }

D. { "aggs": { "unique_users": { "avg": { "field": "user_id" } } } }

Solution

1. Step 1: Identify the aggregation type for unique counts

   The cardinality aggregation is used to count unique values, so the key must be "cardinality".

2. Step 2: Check the field name and structure

   The field is correctly specified as "user_id" inside the cardinality object.

3. Final Answer:

   { "aggs": { "unique_users": { "cardinality": { "field": "user_id" } } } } -> Option A

4. Quick Check:

   Correct syntax uses "cardinality" with "field" [OK]

Hint: Cardinality uses "cardinality" key with "field" [OK]

Common Mistakes:

• Using terms aggregation instead of cardinality
• Using sum or avg aggregations for unique counts
• Incorrect JSON structure or missing braces

3.

Given the following Elasticsearch aggregation query:

{
  "aggs": {
    "unique_countries": {
      "cardinality": {
        "field": "country"
      }
    }
  }
}

If the indexed documents have countries: ["US", "US", "FR", "DE", "FR", "JP"], what is the expected approximate output value for unique_countries?

medium

A. 3

B. 4

C. 5

D. 6

Solution

1. Step 1: Identify unique countries in the list

   The countries are US, FR, DE, JP. Counting unique values gives 4.

2. Step 2: Understand cardinality output

   Cardinality aggregation returns approximate unique counts, which should be 4 here.

3. Final Answer:

   4 -> Option B

4. Quick Check:

   Unique countries count = 4 [OK]

Hint: Count distinct values in the field for cardinality [OK]

Common Mistakes:

• Counting duplicates as unique
• Confusing total documents with unique count
• Expecting exact count always (approximate)

4.

Identify the error in this cardinality aggregation query:

{
  "aggs": {
    "unique_users": {
      "cardinality": {
        "field_name": "user_id"
      }
    }
  }
}

medium

A. Aggregation name cannot be "unique_users"

B. Missing closing brace for "aggs"

C. "field_name" should be "field"

D. Cardinality aggregation requires a "size" parameter

Solution

1. Step 1: Check the field parameter name

   The correct key for specifying the field is "field", not "field_name".

2. Step 2: Verify JSON structure and parameters

   The braces are correctly closed, aggregation name is valid, and "size" is not required for cardinality.

3. Final Answer:

   "field_name" should be "field" -> Option C

4. Quick Check:

   Use "field" key for cardinality field [OK]

Hint: Use "field", not "field_name" in cardinality [OK]

Common Mistakes:

• Using wrong key "field_name" instead of "field"
• Adding unnecessary parameters like "size"
• Misnaming aggregation or missing braces

5.

You want to find the approximate number of unique products sold in each category using Elasticsearch. Which aggregation setup correctly combines terms and cardinality aggregations?

{
  "aggs": {
    "by_category": {
      "terms": { "field": "category" },
      "aggs": {
        "unique_products": {
          "cardinality": { "field": "product_id" }
        }
      }
    }
  }
}

hard

A. This setup groups by category and counts unique products per category

B. This setup counts unique categories per product

C. This setup sums product_id values per category

D. This setup finds average product_id per category

Solution

1. Step 1: Understand the terms aggregation

   "terms" groups documents by unique values in "category" field.

2. Step 2: Understand the nested cardinality aggregation

   Inside each category bucket, cardinality counts unique "product_id" values.

3. Final Answer:

   This setup groups by category and counts unique products per category -> Option A

4. Quick Check:

   Terms + cardinality = unique counts per group [OK]

Hint: Nest cardinality inside terms to count unique per group [OK]

Common Mistakes:

• Confusing aggregation order or nesting
• Using cardinality on category instead of product_id
• Expecting sum or average instead of unique count