createSimulation: Create a data simulation and return train/holdout/validation...

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

Description

Create a data simulation and return train/holdout/validation data sets.

Usage

createSimulation(
  num.samples = 100,
  num.variables = 100,
  pct.imbalance = 0.5,
  pct.signals = 0.1,
  bias = 0.4,
  label = "class",
  sim.type = "mainEffect",
  pct.train = 0.5,
  pct.holdout = 0.5,
  pct.validation = 0,
  save.file = NULL,
  verbose = FALSE
)

Arguments

num.samples	An integer for the number of samples
num.variables	An integer for the number of variables
pct.imbalance	A numeric percentage to indicate proportion of the imbalaced samples. 0 means all controls and 1 mean all cases.
pct.signals	A numeric for proportion of simulated signal variables
bias	A numeric for effect size in simulated signal variables
label	A character vector for the name of the outcome column. class for classification and qtrait for regression
sim.type	A character vector of the type of simulation: mainEffect/interactionErdos/interactionScalefree
pct.train	A numeric percentage of samples to use for traning
pct.holdout	A numeric percentage of samples to use for holdout
pct.validation	A numeric percentage of samples to use for testing
save.file	A filename or NULL indicating whether to save the simulations to file
verbose	A flag indicating whether verbose output be sent to stdout

Value

A list with:

train

traing data set

holdout

holdout data set

validation

validation data set

label

the class label/column name

signal.names

the variable names with simulated signals

elapsed

total elapsed time

See Also

Other simulation: createInteractions(), createMainEffects(), createMixedSimulation(), splitDataset()

Examples

num.variables <- 100
num.samples <- 100
pct.imbalance <- 0.5
pct.signals <- 0.1
bias <- 0.4
label <- "class"
sim.type <- "mainEffect"
sim.data <- createSimulation(num.samples=num.samples,
                             num.variables=num.variables,
                             pct.imbalance=pct.imbalance,
                             pct.signals=pct.signals,
                             bias=bias,
                             label=label,
                             sim.type=sim.type,
                             verbose=FALSE)

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