ComputeMaxInfoGains {MDFS}

Max information gains

Description

Max information gains

Usage

ComputeMaxInfoGains(
  data,
  decision,
  contrast_data = NULL,
  dimensions = 1,
  divisions = 1,
  discretizations = 1,
  seed = NULL,
  range = NULL,
  pc.xi = 0.25,
  return.tuples = FALSE,
  interesting.vars = vector(mode = "integer"),
  require.all.vars = FALSE,
  use.CUDA = FALSE
)

Arguments

data

input data where columns are variables and rows are observations (all numeric)

decision

decision variable as a binary sequence of length equal to number of observations

contrast_data

the contrast counterpart of data, has to have the same number of observations - not supported with CUDA

dimensions

number of dimensions (a positive integer; 5 max)

divisions

number of divisions (from 1 to 15; additionally limited by dimensions if using CUDA)

discretizations

number of discretizations

seed

seed for PRNG used during discretizations (NULL for random)

range

discretization range (from 0.0 to 1.0; NULL selects probable optimal number)

pc.xi

parameter xi used to compute pseudocounts (the default is recommended not to be changed)

return.tuples

whether to return tuples (and relevant discretization number) where max IG was observed (one tuple and relevant discretization number per variable) - not supported with CUDA nor in 1D

interesting.vars

variables for which to check the IGs (none = all) - not supported with CUDA

require.all.vars

boolean whether to require tuple to consist of only interesting.vars

use.CUDA

whether to use CUDA acceleration (must be compiled with CUDA)

Value

A data.frame with the following columns:

• IG – max information gain (of each variable)

• Tuple.1, Tuple.2, ... – corresponding tuple (up to dimensions columns, available only when return.tuples == T)

• Discretization.nr – corresponding discretization number (available only when return.tuples == T)

Additionally attribute named run.params with run parameters is set on the result.

Examples

ComputeMaxInfoGains(madelon$data, madelon$decision, dimensions = 2, divisions = 1,
                    range = 0, seed = 0)