regain: Front end to the reGAIN algorithm.

In insilico/npdro: Nearest-neighbor Projected-Distance Regression Optimized

Front end to the reGAIN algorithm.

Description

Tests number of variables/columns and runs either parallel R or parallel C++.

Usage

regain(
  depVarName,
  labelledDataFrame,
  stdBetas = FALSE,
  absBetas = FALSE,
  numCores = 2,
  verbose = FALSE,
  writeBetas = FALSE,
  regressionFamily = "binomial"
)

Arguments

labelledDataFrame	data.frame with variables in columns and samples in rows; the last column should be labeled 'Class' coded 0 or 1.
stdBetas	logical to use standardized beta coefficients.
absBetas	logical to use absolute value of beta coefficients.
numCores	numeric number of processor cores to use in mclapply
verbose	logical to send verbose messages to stdout.
writeBetas	logical indicating whther to write beta values to separate file.
regressionFamily	string glm regression family name.

Details

regain

Value

regainMatrix matrix of variable by variable regression coefficients.

References

• http://www.nature.com/gene/journal/v11/n8/full/gene201037a.html Genes & Immunity Paper

• https://github.com/hexhead/inbixC++ inbix on github

See Also

Other GAIN functions: dcgain(), dmgain(), fitInteractionModel(), fitMainEffectModel(), gainToSimpleSIF(), getInteractionEffects(), getMainEffects(), regainParallel()

Other inbix synonym functions: dcgain(), getMainEffects(), regainParallel()

Examples

data(testdata10)
rinbixRegain <- regain(testdata10, stdBetas = TRUE, absBetas = TRUE)

insilico/npdro documentation built on July 1, 2023, 2:56 p.m.