data_sources

The data_sources section tells the WBR App where to fetch your daily metric data. You can query databases, reference CSV files, or both.

data_sources is optional

If you upload a CSV directly via the UI or API, that CSV becomes the sole data source and data_sources is ignored. Use data_sources when you want the app to fetch data automatically or merge multiple sources.

Database queries

Each entry under data_sources references a named connection from your connections file, and contains one or more named queries.

setup:
  db_config_url: https://example.com/connections.yaml

data_sources:
  MyProdPostgres:              # Must match a connection name
    main_metrics:              # Descriptive query name
      query: >
        SELECT
          date AS "Date",
          "Impressions", "Clicks", "PageViews"
        FROM wbr_sample_1;

    sales_metrics:
      query: >
        SELECT
          date AS "Date",
          "pct_customer_group_1", "pct_customer_group_2"
        FROM wbr_sample_1;

Every query must return a Date column

Alias the date column as "Date" (quoted, capitalized) in your SQL.

Column namespacing

Non-Date columns are automatically prefixed with the query name:

• main_metrics.Impressions
• main_metrics.Clicks
• sales_metrics.pct_customer_group_1

Reference these namespaced names in your metrics section.

Supported databases

Type	connections.yaml type value	Notes
PostgreSQL	postgres	Default port 5432
Amazon Redshift	redshift	Default port 5439, auto-lowercases columns
Snowflake	snowflake	Auto-handles uppercase DATE → Date
Amazon Athena	athena	Async polling; requires s3_staging_dir

See Connections File for full connection config details and credential management.

CSV files (in config)

To merge CSV data alongside database queries, list CSVs under csv_files:

data_sources:
  MyProdPostgres:
    main_metrics:
      query: >
        SELECT date AS "Date", "Impressions" FROM wbr_sample_1;

  csv_files:
    external_metrics:
      url_or_path: ./data/external_metrics.csv
    remote_data:
      url_or_path: https://example.com/data.csv

CSV columns are namespaced the same way: external_metrics.{column}.

Merging behavior

All sources (queries + CSVs) are outer-joined on Date:

• A date present in one source but not another produces NaN for the missing columns
• Multiple rows per date are allowed — your aggf in metrics handles aggregation

CSV upload vs. config: precedence rules

Scenario	Behavior
CSV uploaded directly (UI or API)	Overrides all YAML data_sources
data_sources defined in YAML, no CSV uploaded	DB queries + config CSVs merged
Both CSV uploaded AND data_sources in YAML	Uploaded CSV wins, data_sources ignored

Tip

Use direct CSV upload for prototyping. Use data_sources for production workflows where data is fetched automatically.