Evaluate Python expressions on pipeline data

Use This When

• Creating conditional logic in pipelines without writing custom components
• Computing derived fields from multiple inputs like ratios, averages, or boolean combinations
• Implementing quick filters or transformations for prototyping before custom builds
• Building complex decision trees by combining multiple data streams

What It Does

• Evaluates Python expressions using asteval interpreter with pipeline inputs as variables
• Maps inputs to variables x0, x1, x2, etc. for use in expression string
• Returns result of expression evaluation which can be any type supported by inputs
• Supports arithmetic, comparisons, boolean logic, and limited Python standard library

Works Best With

• Multiple component outputs → this component → conditional routing or gating
• Data streams → this component → derived metrics or computed features
• Integration with detect-zone-transition or check-object-distance for complex rule evaluation
• Quick prototyping before implementing optimized custom component logic

Caveats

• Expression executed per message; avoid complex computations that impact throughput
• Sandboxed interpreter prevents file I/O and most unsafe operations for security
• Type errors in expressions fail the pipeline; validate expression syntax before deployment
• Not suitable for ML inference or heavy computation; use dedicated components instead
• Lambda expressions are not supported by the evaluator (NotImplementedError); use statements/comprehensions instead
• Builtins like getattr are not available inside expressions
• String .format(...) on literals is blocked by the evaluator; prefer f‑strings inside expressions.