Future plans

[smeznar]

• Add parallelization for quicker parameter estimation
• Add support for different types of expressions
• Create a pipeline for that makes evaluation of SR approaches easy (from data set download and parsing to evaluation in a few simple lines of code)
• Add support for custom error function and additional metrics for performance evaluation on final expressions
• Automatically transform results into a table
• Add baselines for easy comparison
• Improve documentation
• Rework how constants work. Instead of needing to add constant symbols, make them independent, maybe when big C occurs as the first letter

[smeznar]

Try NumExpr and some other optimization algorithms for parameter estimation