probKMA_silhouette_filter: Filter Motifs from probKMA Results Based on Silhouette and...

In funMoDisco: Motif Discovery in Functional Data

Filter Motifs from probKMA Results Based on Silhouette and Size Thresholds

Description

This function filters the motifs identified by the 'probKMA' algorithm based on a threshold for the average silhouette index and a minimum size criterion. Motifs that meet or exceed both the silhouette index threshold and the minimum number of curves (size) are retained. The function returns a cleaned version of the input data, including the filtered motifs, their derivatives, and associated matrices.

Usage

probKMA_silhouette_filter(
  probKMA_results,
  silhouette,
  sil_threshold = 0.5,
  size_threshold = 2
)

Arguments

probKMA_results	A list representing the output of the 'probKMA' function with 'return_options = TRUE'. This output includes the motifs, dissimilarity matrices, and membership matrices required for filtering.
silhouette	A list output from the 'probKMA_silhouette' function, which provides silhouette scores for each motif.
sil_threshold	A numeric value representing the threshold for the average silhouette index. Motifs with a silhouette index greater than or equal to this value will be retained. Default is 0.5.
size_threshold	An integer representing the minimum number of curves (size) in a motif cluster. Motifs with a cluster size greater than or equal to this value will be retained. Default is 2.

Value

A list containing the filtered results:

V0_clean	Filtered motifs.
V1_clean	Derivatives of the filtered motifs.
D	Dissimilarity matrix.
D_clean	Filtered dissimilarity matrix after cleaning motifs.
P	Membership matrix.
P_clean	Filtered membership matrix after cleaning motifs.
c	Filtered minimum motif lengths.
K	Vector containing the number of motifs repeated by the filtered motifs.

funMoDisco documentation built on April 16, 2025, 1:10 a.m.