scorepyo.ranking#

Classes for binary features rankers

Classes

BordaRank(list_ranker, **kwargs)

Based on a list of Ranker, computes the Borda rank of each binary feature based on all rankers.

CumulativeMetric(metric, ranker, **kwargs)

Binary features ranker based on computing a metric on a growing number of binary features.

DiverseLogOddsDensity([rank_diversity])

Binary features ranker based on logodds contribution of binary feature, and density of the binary feature (number of samples with a positive value on the binary feature)

FasterRiskRank([parent_size, child_size, ...])

Binary features ranker based on FasterRisk, another risk score model library.

LarsPathRank(**kwargs)

Binary feature ranker based on LARS lasso path.

LassoPathRank(**kwargs)

Binary feature ranker based on lasso path.

LogOddsDensity(**kwargs)

Binary features ranker based on logodds contribution of binary feature, and reduced importance of density of the binary feature (number of samples with a positive value on the binary feature)

OMPRank(**kwargs)

Binary feature ranker based on Orthogonal Matching Pursuit.

Ranker()

Base class for binary features ranker