sure_independence_screen: Sure independence screening function

In robertzk/syberiaMungebits: Common mungebits used with Syberia projects.

Description

Sure independence screening function

Usage

sure_independence_screen(dataframe, ..., exclude = character(0),
  max_levels = 100, discretizer_params = list())

Arguments

dataframe	a data.frame. Run sure independence screening on this dataframe.
...	further SIS options. See also sure_independence_screening in the statsUtils package.
exclude	character. A vector of variables to never remove.
discretizer_params	list. A list of parameters to pass to the discretizer that discretizes all non-categorical variables before sure independence screening. By default, these are granularity = N, upper_count_bound = NULL, lower_count_bound = 1, where N is the rounded mean of the number of levels in each existent factor variable (i.e. if some were pre-discretized), or 3 if there are less than 5 such variables. If any are left undiscretized, these will not be screened.

robertzk/syberiaMungebits documentation built on May 25, 2017, 6:40 a.m.