gpd-parameter-sweep

You can automate a parameter sweep by defining parameter ranges and letting the system evaluate independent points in parallel. It collects results at each point, aggregating them into a structured dataset for downstream analysis.

Adaptive refinement in parameter space exploration automatically zooms in on regions exhibiting rapid changes, extrema, or interesting transitions. This targets computational resources toward mapping critical areas rather than uniformly sampling the entire range.

Yes, you can run parallel parameter sweeps for two variables simultaneously. The system supports varying one or two parameters across defined ranges, executing independent wave-based evaluations to accelerate discovery across the defined grid.

The parameter sweep workflow tracks successes and failures for each data point. Every evaluated sweep point reports its success status, failure state, and associated metadata, which is then automatically aggregated into the final structured dataset.

The primary prerequisite for an automated parameter sweep is that each point must use the exact same computation method across the defined ranges. This consistency enables parallel evaluation and the subsequent automatic aggregation into a structured dataset.

The best way to map a parameter space and identify extrema is using an automated parallel sweep with adaptive refinement. This approach systematically explores the ranges, zooms into transitions, and produces a concise summary report ready for publication.