Use sweep setup, sweep detail, and optimizer surfaces when strategy selection and allocation need deeper tooling.
Use this guide when you need to move from one-off strategy validation into broader parameter search and allocation design across the signed-in strategy tooling.
Sweeps are for parameter search and candidate ranking. Optimizer is for portfolio construction once the candidate universe is clearer.
Treat sweep detail as an evidence surface with leaderboard, Pareto, statistics, and compare actions rather than just a run log.
Keep the handoff into compare, backtests, and allocation explicit so strategy search and position design do not drift apart.
These are the deeper strategy-tooling surfaces behind strategy exploration and allocation design.
Define universe, strategy, parameter grid, density, and ranking policy before you launch a sweep.
Review prior sweep runs and reuse their context instead of relaunching without a history check.
Inspect progress, leaderboard, Pareto rows, statistics, memo output, and selected-run comparison for a sweep.
Create and tune allocation profiles, solver settings, and constraint policies before you solve anything.
Select the universe, seed expected returns, tune bounds, and inspect allocation diagnostics plus result rows.
Use this sequence to keep strategy search and allocation design connected but distinct.
Start by fixing the watchlist, instrument, interval, parameter grid, and ranking policy so the search space is explicit.
Use leaderboard, Pareto, statistics, and memo surfaces to understand why the best-looking runs are winning.
Move selected runs into compare or deeper backtest review before you treat the sweep winner as final.
Use optimizer profiles, solver constraints, and allocation diagnostics after the candidate universe is already credible.
Do not treat sweep ranking as enough by itself. Use sweep detail and compare to validate why a candidate is winning.
Keep parameter-grid assumptions explicit so later sweeps are comparable rather than cosmetically different.
Use optimizer profiles and bounds as governance controls, not just numeric tuning knobs.
Separate strategy selection from allocation design so a weak strategy is not disguised by a good optimizer outcome.
Jump into the sweeps surface to define the search space and inspect recent sweep runs.
Open the optimizer when the candidate set is ready for explicit profile, constraint, and allocation design.
Return to run-detail validation when a sweep winner or optimizer output still needs deeper backtest evidence.
Mar 24, 2026
Report unclear guidance, stale contracts, missing coverage, or broken docs UI on this page.
Open feedback issueJump to the section you need without losing your place.
Mar 24, 2026
Report unclear guidance, stale contracts, missing coverage, or broken docs UI on this page.
Open feedback issue