Row Count Check

The row_count check validates the number of rows in a dataset. This is one of the most basic and commonly used data quality checks.

Configuration

Parameter	Required	Description
name	Yes	Unique name for the check
dataset	Yes	Table name or SQL query
type	Yes	Must be row_count
condition	Yes	Comparison operator
threshold	Yes	Expected row count value
dimensions	No	Group by columns
filter	No	WHERE clause conditions

Examples

Basic Row Count Check

- name: orders_minimum_rows
  dataset: orders
  type: row_count
  condition: gt
  threshold: 0

This check ensures the orders table has more than 0 rows.

Row Count with Filter

- name: active_users_count
  dataset: users
  type: row_count
  condition: ge
  threshold: 100
  filter: status = 'active'

This check validates that there are at least 100 active users.

Row Count by Dimensions

- name: orders_by_region
  dataset: orders
  type: row_count
  dimensions: [region, status]
  condition: gt
  threshold: 10

This check ensures each region/status combination has more than 10 orders.

Row Count with Time Dimension

- name: daily_orders_count
  dataset: orders
  type: row_count
  condition: gt
  threshold: 50
  time_dimension:
    name: created_at
    granularity: day

This check validates that each day has more than 50 orders.

Use Cases

• Data Freshness: Ensure new data is being loaded
• Completeness: Verify expected data volume
• Business Rules: Validate minimum activity levels
• Monitoring: Track growth or decline in data volume

Generated SQL

The row count check generates SQL similar to:

SELECT COUNT(*) 
FROM orders 
WHERE status = 'active'

With dimensions:

SELECT region, status, COUNT(*) 
FROM orders 
GROUP BY region, status