StockSharp's public API now includes a dedicated set of components for strategy optimization. Full parameter sweeps and genetic search are available, while the sample application also demonstrates random sampling of parameter combinations. Optimization ranges are validated before a run begins, helping detect invalid steps or incompatible boundaries earlier.
OptimizationBatchManager manages parallel iterations without exceeding the configured batch size. Users can limit the number of simultaneous tests, the total number of iterations, and the number of processed messages. Genetic optimization provides settings for population size, generation count, stagnation criteria, mutation, crossover, and selection.
The objective function is defined as a formula based on strategy statistics, including profit, drawdown, trade count, latency, order errors, and other metrics. The ready-to-run sample displays the progress and P&L of each run and demonstrates the complete optimization setup and shutdown workflow.
Comments (0)
Login or Create account, Log in or register to leave a comment
No comments yet. Be the first!