Skip to main content

Profiling Configuration

This guide explains how to configure profiling defaults and dataset-level overrides.

Where to Configure​

  • Global defaults: config.yml under profiling
  • Dataset-specific overrides: datasets/{table}.yml (recommended)
  • Schema/database defaults: {schema}_schema.yml, {database}_database.yml

Example global defaults:

profiling:
  default_sample_ratio: 1.0
  compute_histograms: true
  max_distinct_values: 1000

Example dataset override:

# datasets/orders.yml
database: warehouse
schema: sales
table: orders

profiling:
  sampling:
    enabled: true
    fraction: 0.1
  partition:
    key: created_at
    strategy: latest
  columns:
    - name: order_total
      metrics: [mean, max, min]

Precedence​

  1. Table file overrides
  2. Schema file overrides
  3. Database file overrides
  4. Global defaults

UI Workflow​

  • Use the Profiling page for global defaults.
  • Use Datasets → Dataset Detail for per-dataset and column overrides.
  • Preview merged configs and precedence in the Datasets page before saving.

Validation​

  • CLI: baselinr validate-config --config config.yml
  • UI: Datasets page validation and merged preview
  • DATASET_CONFIGURATION.md
  • PARTITION_SAMPLING.md
  • PROFILING_ENRICHMENT.md