hierarchical_groups: Creates hierarchically structured candidate groups based on a...

In amspector100/blipr: A Bayesian Linear Program for resolution-adaptive signal detection

Creates hierarchically structured candidate groups based on a distance matrix.

Description

Creates hierarchically structured candidate groups based on a distance matrix.

Usage

hierarchical_groups(
  samples,
  dist_matrix = NULL,
  X = NULL,
  max_pep = 1,
  max_size = 25,
  filter_sequential = FALSE
)

Arguments

samples	(N,p)-shaped array of posterior samples where a nonzero value indicates the presence of a signal.
dist_matrix	A distance matrix corresponding to distances between locations, used for hierarchical clustering.
X	The design matrix in regression problems, which will be used to create dist_matrix if dist_matrix is not provided.
max_pep	The maximum posterior error probability (PEP) allowed in a candidate group. Default is 1.
max_size	maximum allowable size for each group.
filter_sequential	If TRUE, ignore sequential groups of variables to avoid duplication.

amspector100/blipr documentation built on June 22, 2022, 2:43 a.m.