After a successful run, you can access detailed metrics and validation results from the run.metadata object. This example shows how to parse this object. To make this example runnable, we’ll create a mock LumeRun object that simulates the result of a completed pipeline. In a real workflow, this run object would be the output of lume.run(...) followed by run.wait(). See the Advanced Topics page for the full, detailed schema of the metadata object. 
class MockLumeRun:
    """A mock LumeRun object for demonstration purposes."""
    def __init__(self, status, metadata):
        self.status = status
        self.metadata = metadata

# Create a mock run object with sample data
run = MockLumeRun(
    status="PARTIAL_FAILED",
    metadata={
        "pipeline": {
            "duration_seconds": 12.345,
            "sync_source_seconds": 4.1,
            "transform_seconds": 7.2,
            "sync_target_seconds": 1.045
        },
        "results": {
            "input_rows": 1000,
            "mapped_rows": 990,
            "rejected_rows": 10
        },
        "validation": {
            "tests_executed": 2970,
            "error_rate": 0.01,
            "top_errors": [
                {"field": "email", "error_code": "INVALID_FORMAT", "count": 8},
                {"field": "postal_code", "error_code": "VALUE_TOO_SHORT", "count": 2}
            ]
        }
    }
)

# --- Start of parsing logic ---
# You can copy-paste everything below this line to process a real run object.

if run.status in ["SUCCEEDED", "PARTIAL_FAILED"]:
    
    metadata = run.metadata
    
    # --- Performance Metrics ---
    pipeline_metrics = metadata.get("pipeline", {})
    print("\n--- Pipeline Performance ---")
    print(f"Total Duration: {pipeline_metrics.get('duration_seconds', 0):.2f}s")
    
    # --- Row Counts ---
    result_metrics = metadata.get("results", {})
    print("\n--- Row Counts ---")
    print(f"Input Rows: {result_metrics.get('input_rows', 'N/A')}")
    print(f"Mapped Rows: {result_metrics.get('mapped_rows', 'N/A')}")
    print(f"Rejected Rows: {result_metrics.get('rejected_rows', 'N/A')}")

    # --- Validation Summary ---
    validation_metrics = metadata.get("validation", {})
    print("\n--- Validation Summary ---")
    print(f"Total Tests Executed: {validation_metrics.get('tests_executed', 'N/A')}")
    print(f"Error Rate: {validation_metrics.get('error_rate', 0):.2%}")

    # --- Top Errors ---
    top_errors = validation_metrics.get("top_errors", [])
    if top_errors:
        print("\n--- Top 5 Validation Errors ---")
        for error in top_errors[:5]:
            print(f"- Field: {error.get('field')}, Error: {error.get('error_code')}, Count: {error.get('count')}")