Agregacje kubełkowe

Welcome to Bucket Aggregations! In Elasticsearch, aggregations allow you to analyze your data.

Bucket aggregations specifically group documents into sets, or 'buckets', based on field values. Think of it like the GROUP BY clause in SQL.

• Categorize data: Group products by brand.
• Faceted search: Show counts for different filters.
• Analyze trends: See activity per day or month.

The terms aggregation is one of the most common bucket aggregations. It finds the top unique values for a specific field and then counts how many documents fall into each unique value's bucket.

It's incredibly useful for seeing the distribution of categorical data, like product categories, user roles, or country names.

Let's use a terms aggregation to find the top 5 product categories and the count of products in each. We use .keyword for exact matches on text fields.

The range aggregation allows you to define custom ranges for numeric or date fields. Documents whose field values fall within a defined range are grouped into that bucket.

This is perfect for creating price tiers (e.g., $0-10, $10-50, $50+) or age groups (e.g., 0-18, 19-65, 65+).

Here, we define three price ranges: products under $10, between $10 and $50, and over $50. The size: 0 means we only want aggregation results, not actual search hits.

When working with time-series data, the date_histogram aggregation is your best friend. It buckets documents into fixed time intervals like minutes, hours, days, or months.

You specify an interval (e.g., 'day', 'month') and Elasticsearch automatically creates buckets for each period, even if no documents exist for a particular period.

This aggregation groups sales orders by month, allowing you to easily track monthly sales performance. We assume an order_date field of type date.

The true power of bucket aggregations comes from nesting them. You can place one bucket aggregation inside another to create hierarchical groupings.

For example, you might want to see product categories, and then within each category, the price ranges of products. This enables deep, multi-dimensional analysis.

Bucket aggregations are fundamental for many real-world applications:

• Faceted Search: Allowing users to filter search results by category, brand, price range, etc.
• Data Exploration: Discovering patterns and distributions in your data.
• Dashboards: Building visualizations like bar charts (e.g., sales per month, products per category).
• Reporting: Generating summary reports based on grouped data.

You are analyzing user feedback and want to count how many reviews were submitted each day over the past week. Which aggregation type is most suitable?

Great job! You've learned about the power of Bucket Aggregations in Elasticsearch.

• They group documents into logical 'buckets'.
• terms: Groups by unique field values.
• range: Groups by custom numeric or date ranges.
• date_histogram: Groups by fixed time intervals.
• You can nest them for multi-level analysis.

Next, we'll explore Metric Aggregations, which perform calculations (like sum, avg, min, max) on the data within these buckets!