Systematically evaluating hypotheses
Proximity hypothesis
The score_by_proximity_y function evaluates how successfully the proximity hypothesis predicts the layout of scholia for a single page by comparing the model's placement of scholia with their actual recorded positions.
It has optional parameters that allow you to limit the analysis to a specific set of scholia (default: msA, the main scholia of the Venetus A manuscript), and to set a threshhold distance for how close a placement constitutes a successful predicted layout. The threshhold is scaled from 0.0 to 1.0. A threshhold of 1.0 matches anything on the page; a threshhold of 0.0 matches nothing. The default value of 0.1 considers the model's placement successful when its y value falls within 10% of the actual y value. Compare the following examples
score_proximity = score_by_proximity_y(pg; threshhold = 0.1)PageScore(urn:cite2:hmt:msA.v1:112r, 3, 9)success_rate(score_proximity)0.25score_15 = score_by_proximity_y(pg; threshhold = 0.15)PageScore(urn:cite2:hmt:msA.v1:112r, 7, 5)success_rate(score_15)0.583Zoned layout hypothesis
The score_by_zones function evaluates how successfully the zoned page layout hypothesis predicts the layout of scholia for a single page by comparing the model's placement of scholia with their actual recorded positions.
score_zones = score_by_zones(pg)PageScore(urn:cite2:hmt:msA.v1:112r, 6, 6)success_rate(score_zones)0.5Caveats
Missing data for even a single scholion on a page can skew the model's results.