Configuration

Project Configuration

Configure your project with fyrnheim.yaml at the project root:

entities_dir: entities
data_dir: data
output_dir: generated
backend: duckdb
backend_config:
  db_path: my_project.duckdb

# Push results to a separate output backend after fyr run
output_backend: clickhouse
output_config:
  host: localhost
  port: "8123"
  database: default
  user: default
  password: ""

Configuration Options

Option	Description	Default
entities_dir	Directory containing entity Python files	entities
data_dir	Directory containing local data files (parquet, CSV)	data
output_dir	Directory where generated transform code is written	generated
backend	Transformation backend (duckdb, bigquery)	duckdb
backend_config	Backend-specific configuration (see below)	{}
output_backend	Optional output backend for pushing results	none
output_config	Output backend-specific configuration	{}

Backend Configuration

DuckDB:

backend: duckdb
backend_config:
  db_path: my_project.duckdb

BigQuery:

backend: bigquery
backend_config:
  project: my-gcp-project
  dataset: analytics

CLI Flags

All settings can be overridden via CLI flags. For example:

fyr run --backend bigquery

This runs on BigQuery regardless of what fyrnheim.yaml says.

Use fyr --help to see all available CLI options:

fyr --help
fyr generate --help
fyr run --help

Production Deployment

A typical production pattern:

1. Extract raw data with DLT (or any EL tool) into parquet files or a warehouse
2. Transform with Fyrnheim: fyr run --backend bigquery (or duckdb for local)
3. Push results to an output backend for serving (ClickHouse, Postgres, etc.)

Configure the output backend in fyrnheim.yaml:

backend: duckdb           # transform backend
output_backend: clickhouse # push dim/analytics tables here after run
output_config:
  host: ch.example.com
  port: "8123"
  database: analytics

This separation lets you develop locally on DuckDB while pushing production results to a fast query engine.