Overview

Beyond basic field and mapping management, Lytics provides advanced schema capabilities for analyzing field usage, visualizing relationships between fields, and testing schema expressions.

Field Analytics

Field Information

GET /v2/schema/{table}/fieldinfo

Returns detailed metadata about field values, including usage statistics, sample values, and cardinality. This is useful for understanding the distribution and quality of data across your schema fields.

Query Parameters

Response

{
  "fields": [
    {
      "name": "email",
      "count": 50000,
      "values": [
        {"value": "[email protected]", "count": 1},
        {"value": "[email protected]", "count": 1}
      ],
      "cardinality": 48500,
      "first_seen": "2023-06-01T00:00:00Z",
      "last_seen": "2024-01-15T10:30:00Z"
    }
  ]
}

📘

Field information is cached for 5 minutes. Repeated requests within this window return cached results.

Field Names

GET /v2/schema/{table}/fieldnames

Returns a list of all field names defined in the specified table. This is a lightweight alternative to fetching the full schema when you only need field names.

Schema Visualization

Field Relationship Network

GET /v2/schema/{table}/network

Returns a graph of relationships between fields in the schema. This network visualization shows how fields are connected through identity resolution, mappings, and data flow.

The response includes nodes (fields) and edges (relationships) that can be rendered as a network graph. Results are cached for performance.

Expression Evaluation

Evaluate Schema Expression

POST /v2/schema/{table}/evaluate

Evaluates test events against a schema and returns the profiles they would produce, without writing anything. By default it evaluates the current draft. Pass version_id to evaluate a specific published version, or patch_id to evaluate a draft merged onto the live schema, applied the same way publishing applies it, so the result matches what you would see after publishing. Events that resolve no identity or produce no change are skipped; the request fails only when nothing resolved at all.

This is useful for:

• Verifying that a mapping expression produces the expected output
• Testing filter conditions before using them in stream routing rules
• Debugging complex LQL expressions