5. Analyzing and Using Optimization Results
5.1. Result Analysis
Sort the result table by Sharpe ratio, drawdown, and other columns to compare stability and risk-return profiles across parameter sets.
If supported, plot parameter–objective relationships to inspect sensitivity and local optima.
5.2. Objectives and Constraints (with Brief Explanations)
Optimization objective: Currently supported goals such as maximizing Sharpe ratio, minimizing drawdown, or maximizing return; see configuration or API docs.
Constraints: If any (e.g., max drawdown limit or minimum trade count), set them in configuration; only feasible solutions participate in ranking.
5.3. Apply Best Parameters to the Strategy
Use op.set_parameter(stg_id, pars=best_pars) to apply optimized parameters to the Operator; the same Operator can then be used for backtesting (mode=1) or live trading without manual code changes.
5.4. Caveats
Overfitting: In-sample optimized parameters may underperform out of sample; use out-of-sample validation or rolling optimization.
In-sample vs out-of-sample: Use different date ranges for optimization and validation.
Compute time: Larger
opti_sample_countor iteration counts take longer; balance accuracy against runtime.