compute_minimal_intervals: Computes the set of minimal intervals as described in...

In MSinference: Multiscale Inference for Nonparametric Time Trend(s)

Computes the set of minimal intervals as described in Duembgen (2002)

Description

Given a set of intervals, this function computes the corresponding subset of minimal intervals which are defined as follows. For a given set of intervals \mathcal{K}, all intervals \mathcal{I}_k \in \mathcal{K} such that \mathcal{K} does not contain a proper subset of \mathcal{I}_k are called minimal.

This function is needed for illustrative purposes. The set of all the intervals where our test rejects the null hypothesis may be quite large, hence, we would like to focus our attention on the smaller subset, for which we are still able to make simultaneous confidence intervals. This subset is the subset of minimal intervals, and it helps us to to precisely locate the intervals of further interest.

More details can be found in Duembgen (2002) and Khismatullina and Vogt (2019, 2020)

Usage

compute_minimal_intervals(dataset)

Arguments

dataset

Set of the intervals. It needs to contain the following columns: "startpoint" - left end of the interval; "endpoint" - right end of the interval.

Value

Subset of minimal intervals

Examples

startpoint   <- c(0, 0.5, 1)
endpoint     <- c(2, 2, 2)
dataset      <- data.frame(startpoint, endpoint)
minimal_ints <- compute_minimal_intervals(dataset)

MSinference documentation built on Sept. 11, 2024, 5:38 p.m.