RaSEn (version 3.0.0)

RaRank: Rank the features by selected percentages provided by the output from RaScreen.

Description

Rank the features by selected percentages provided by the output from RaScreen.

Usage

RaRank(object, selected.num = "all positive", iteration = object$iteration)

Arguments

object

output from RaScreen.

selected.num

the number of selected variables. User can either choose from the following popular options or input an positive integer no larger than the dimension.

• 'all positive': the number of variables with positive selected percentage.

• 'D': floor(D), where D is the maximum of ramdom subspace size.

• '1.5D': floor(1.5D).

• '2D': floor(2D).

• '3D': floor(3D).

• 'n/logn': floor(n/logn), where n is the sample size.

• '1.5n/logn': floor(1.5n/logn).

• '2n/logn': floor(2n/logn).

• '3n/logn': floor(3n/logn).

• 'n-1': the sample size n - 1.

• 'p': the dimension p.

iteration

indicates results from which iteration to use. It should be an positive integer. Default = the maximal interation round used by the output from RaScreen.

Value

Selected variables (indexes).

References

Tian, Y. and Feng, Y., 2021(a). RaSE: A variable screening framework via random subspace ensembles. Journal of the American Statistical Association, (just-accepted), pp.1-30.

Examples

# NOT RUN {
set.seed(0, kind = "L'Ecuyer-CMRG")
train.data <- RaModel("screening", 1, n = 100, p = 100)
xtrain <- train.data$x
ytrain <- train.data$y

# test RaSE screening with linear regression model and BIC
fit <- RaScreen(xtrain, ytrain, B1 = 100, B2 = 50, iteration = 0, model = 'lm',
cores = 2, criterion = 'bic')

# Select floor(n/logn) variables
RaRank(fit, selected.num = "n/logn")
# }