paperSimWorkflow: Workflow for running one simulation of the Bioinformatics...

In insilico/privateEC: Privacy Preserving Evaporative Cooling Feature Selection and Classification with Relief-F and Random Forests

Description

Creates one simulation of train/holdout/validation data sets, then runs the four algorithms on that data. Returns a data frame of run results for each.

Usage

paperSimWorkflow(
  n.samples = 100,
  n.variables = 100,
  pct.signals = 0.1,
  update.freq = 10,
  verbose = FALSE
)

Arguments

n.samples	An integer for the number of samples
n.variables	An integer for the number of variables
pct.signals	A numeric for the significant variable bias
update.freq	A integer for the number of steps before update
verbose	A flag indicating whether verbose output be sent to stdout

Value

A list containing:

run.results

data frame of run results of each sim type

elapsed

total elapsed time

Note

Default parameter values match those from the Bioinformatics paper.

See Also

The workflow consists of the sequence: createSimulation privateEC originalThresholdout privateRF standardRF and compileAndSaveAllResults. A comparison analysis with real data (fMRI) is in paperRealWorkflow.

Other workflows: compileAndSaveAllResults(), paperRealWorkflow(), plotRunResults()

Examples

  num.samples <- 100
  num.variables <- 100
  pct.signals <- 0.1
  upd.frq <- 0.1 * num.variables
  one.step.result <- paperSimWorkflow(n.samples = num.samples,
                                      n.variables = num.variables,
                                      pct.signals = pct.signals,
                                      update.freq = upd.frq,
                                      verbose = FALSE)

insilico/privateEC documentation built on May 22, 2020, 5:12 p.m.