Enhance Genetic Algorithm Optimization with User-Configurable "Population Size" and "Generations" Parameters, and API Customization Options

I would like to suggest two enhancements to cTrader’s optimization capabilities, particularly for the Genetic Algorithm (GA) method, to improve flexibility and control for advanced users developing cBots.

1. Expose "Population Size" and "Generations" Parameters for User Configuration:
 

Currently, cTrader’s Genetic Algorithm for cBot optimization uses "Population Size" and "Generations" internally, but these parameters are not user-configurable in the interface (as observed in cTrader version 5.1.14). I propose exposing these settings in the Optimization interface, allowing users to adjust them (e.g., Population Size: 50–200, Generations: 20–50). This would enable fine-tuning of the GA process to balance speed and accuracy, especially for complex strategies with multiple parameters, reducing the risk of premature convergence or suboptimal results.
 

2. Provide Customization Options for GA via cTrader API or Other Tools:
 

In addition to exposing GA parameters in the UI, I suggest adding support for customizing the GA process (e.g., mutation rate, crossover rate, fitness function adjustments) through the cTrader API or a dedicated plugin/extension system. This would allow advanced users to implement custom GA logic tailored to specific trading strategies, potentially integrating external tools or libraries for more sophisticated optimization. For example, enabling developers to define custom fitness functions or control GA behavior programmatically would greatly enhance the platform’s flexibility for algorithmic trading.
 

These enhancements would benefit advanced cBot developers by offering greater control over optimization, reducing processing time for complex strategies, and improving the accuracy of parameter tuning. I believe this would align with cTrader’s goal of providing powerful tools for algo trading while maintaining user-friendliness.

Thank you for considering this suggestion, and I look forward to any feedback or updates on plans to implement such features in future versions.