Ranking

Orders scored assets and selects a subset of the connected universe.

When to use

Use it when the Portfolio graph should choose candidates by a connected score.

Example

Rank the filtered universe by a momentum score and pass the selected assets to Weight Calculator.

Try it here

Explore Ranking

Inspect an intentionally incomplete Ranking example and its required Scores input.

Exercise steps

1. Change the direction and compare Top N with percentage selection.
2. Inspect the missing Scores input. It requires a metric output explicitly typed as Ranking.

Start lesson

Select from scored assets

Ranking needs both an Asset Universe and a compatible Scores ranking input; it does not calculate the scores itself. For example, imagine three candidate assets with scores 0.9, 0.6, and no score. Set Direction to Higher is better, Selection to Top N, Count to 2, and Null policy to Exclude. The two scored assets are selected, in score order, while the unscored asset is excluded. This is an illustrative input, not a promise that a specific metric emits those values.

Switching to Lower is better reverses the preferred score direction. A Bottom N or percentage selection asks a different question and may change the asset count. Inspect the Selected universe before Weight Calculator; if it is empty, a later target-weight or rebalance stage cannot act as though two assets were selected.

Common problems

Scores must correspond to the connected universe. Review ascending or descending direction and selection limits.

Next steps

Review the generated Parameters, Outcomes, and Data Ports on this page, then open the related guides for the complete workflow.