# Warehouse-native Amplitude: Best Practices

> For AI agents: a documentation index is available at [/docs/llms.txt](/docs/llms.txt). Append `.md` to any page URL for markdown, or send `Accept: text/markdown`.

> **Warning:**
>
> Warehouse Native is a legacy feature and isn't available to new customers.

Warehouse-native Amplitude lets you bring your own models to your analyses. To get the most from your data, consider these best practices:

## Clustering key

- Choose clustering key columns based on the query patterns and filtering conditions in your analytics workload.

  - For example, for event (fact) tables, cluster on event time with the `LINEAR()` function.
- Avoid columns with high cardinality as clustering keys. High-cardinality columns reduce storage efficiency and query performance.
- Use composite clustering keys with multiple columns common in join operations or filtering.

## Schema format

- Use a star schema or Snowflake schema to optimize query performance and simplify data analysis.

  - Star schema: a central fact table that links to dimension tables, suited to simpler queries and faster aggregations.
  - Snowflake schema: a normalized version of the star schema that minimizes data redundancy and improves data integrity at the cost of more complex queries.

## Partition and clustering

Partition large tables to reduce the amount of data scanned and improve query performance.