Running an experiment

The Run step is a live dashboard. Once you start an experiment, it shows progress as it goes, so you always know how far along it is and how it is doing.

Progress at a glance

Four cards along the top track points tried, runs completed, time elapsed, and estimated time remaining, with a progress bar underneath. Because the search knows its budget, the estimate is honest about how much is left.

Best result so far

The chart plots each configuration as it finishes, in the order it ran, against your primary response. A line traces the best result found so far, so you can watch it climb and then level off as the search settles. Feasible points are filled; points that broke a limit show as hollow rings; a dashed line marks a target if you set one.

What is moving the needle

Beside the chart, a panel ranks the factors by how much they are affecting the result right now, longest bar first. This tells you which settings are doing the work while the run is still going, so a study that turns out to hinge on one factor is obvious early.

Every option gets a fair test

By default, each configuration is tested against the same random conditions, replication for replication. This paired comparison means the differences you see come from the settings you changed, not from one configuration happening to get an easier run. It makes the comparison much sharper at no cost to you.

Pause, stop, and finish

You are never locked in. Pause to hold the run, or stop and keep results to end early and keep everything gathered so far, which is safe because every finished point is already recorded. The search also stops on its own when results stop improving. When it is done, a summary appears with Export results and View Results, which takes you to the Results step.