ComputeLocalErrorMetafeature: Computes the importance as the median absolute error for...

In ncats/MultiOmicsGraphPrediction: Network-Based Ensemble Prediction from IntLIM Predictors

Computes the importance as the median absolute error for local predictors (i.e. the predictions for k nearest neighbors of each sample).

Description

Computes the importance as the median absolute error for local predictors (i.e. the predictions for k nearest neighbors of each sample).

Usage

ComputeLocalErrorMetafeature(
  predictions,
  true,
  k,
  inputData,
  eigStep = 10,
  alphaMin = 0,
  alphaMax = 1,
  alphaStep = 0.1
)

Arguments

predictions	Prediction data frame, where rows are samples, and columns are predictors.
true	Named vector of the true outcome values.
k	The number of nearest neighbors to consider.
inputData	The input data read in using the function IntLIM function ReadData.
eigStep	The number of eigenvectors to step by during the evaluation. Note that this must be less than the number of samples. Default = 10.
alphaMin	The lowest value of alpha to investigate. Default = 0.
alphaMax	The highest value of alpha to investigate. Default = 1.
alphaStep	The value of alpha to step by during the evalutation. Default = 0.1.

Value

A data frame with the importance metric for each sample and each predictor.

ncats/MultiOmicsGraphPrediction documentation built on Aug. 23, 2023, 9:19 a.m.