GetMetaFeatures: Computes the metafeatures for each sample and model.

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

Computes the metafeatures for each sample and model.

Description

Computes the metafeatures for each sample and model.

Usage

GetMetaFeatures(
  predictions,
  inputData,
  metaFeatureList = c("pdf", "localerr", "globalerr", "pathway", "reaction",
    "interactionpval", "interactioncoef", "analytecoef", "equality"),
  modelStats = "",
  k = k,
  eigStep = 10,
  alphaMin = 0,
  alphaMax = 1,
  alphaStep = 0.1,
  stype = "",
  colIdInd = "",
  colIdOut = "",
  modelsToConsider
)

Arguments

predictions	Prediction data frame, where rows are samples, and columns are predictors.
inputData	The input data read in using the function IntLIM function ReadData.
metaFeatureList	A list of the valid metrics to include. Valid metrics are "pdf", "localerr", "globalerr", "pathway", "reaction", "interactionpval", "interactioncoef", and "analytecoef". Additionally, use "equality" to add equal weight to all predictors.
modelStats	A data frame that includes the interaction p-values and interaction coefficients for each pair (such as the one output by IntLIM's ProcessResults function)
k	The number of nearest neighbors to consider in localerr.
eigStep	The number of eigenvectors to step by during the evaluation in localerr. Note that this must be less than the number of samples in localerr. Default = 10.
alphaMin	The lowest value of alpha to investigate in localerr. Default = 0.
alphaMax	The highest value of alpha to investigate in localerr. Default = 1.
alphaStep	The value of alpha to step by during the evaluation in localerr. Default = 0.1.
stype	Phenotype or outcome to use in models.
colIdInd	The ID of the column that has the analyte ID's for the independent variable. If blank, then the existing ID's are used.
colIdOut	The ID of the column that has the analyte ID's for the outcome variable. If blank, then the existing ID's are used.
modelsToConsider	A list of models for which to calculate meta-features.

Value

A list of data frames (one for each sample) with predictor importance measured according to the listed criteria (one column per metric, one row per predictor).

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