EpistasisKatz: Rank variables by EpistasisKatz algorithm.
In insilico/glmSTIR: Nearest-neighbor Projected-Distance Regression

EpistasisKatz

R Documentation

Rank variables by EpistasisKatz algorithm.

This function can be use as original Katz algorithm or as EpistasisKatz that incorporates prior knowledge.

EpistasisKatz(A = NULL, alpha = NULL, beta = NULL, magnitude.sort = T)

`A`	`matrix` network of features either in adjacency or interaction format.
`alpha`	`numeric` a vector with numeric values.
`beta`	`numeric` either a vector of constant values or prior knowledge.
`magnitude.sort`	default (T) is to rank variable scores by their magnitude so negative effects are also ranked. Using False will force negative effects to be ranked last.