caps: Class for clustering with amplitude and phase separation

In fdacluster: Joint Clustering and Alignment of Functional Data

Class for clustering with amplitude and phase separation

Description

The k-means algorithm with joint amplitude and phase separation produces a number of outputs. This class is meant to encapsulate them into a single object for providing dedicated S3 methods for e.g. plotting, summarizing, etc. The name of the class stems from Clustering with Amplitude and Phase Separation.

Usage

as_caps(x)

is_caps(x)

Arguments

x	A list coercible into an object of class caps.

Details

An object of class caps is a list with the following components:

• original_curves: A numeric matrix of shape N \times L \times M storing a sample with the N L-dimensional original curves observed on grids of size M.

• original_grids: A numeric matrix of size N \times M storing the grids of size M on which original curves are evaluated;

• aligned_grids: A numeric matrix of size N \times M storing the grids of size M on which original curves must be evaluated to be aligned;

• center_curves: A numeric matrix of shape K \times L \times M storing the K centers which are L-dimensional curves observed on a grid of size M;

• center_grids: A numeric matrix of size K \times M storing the grids of size M on which center curves are evaluated;

• warpings: A numeric matrix of shape N \times M storing the estimated warping functions for each of the N curves evaluated on the within-cluster common grids of size M;

• n_clusters: An integer value storing the number of clusters;

• memberships: An integer vector of length N storing the cluster ID which each curve belongs to;

• distances_to_center: A numeric vector of length N storing the distance of each curve to the center of its cluster;

• silhouettes: A numeric vector of length N storing the silhouette values of each observation;

• amplitude_variation: A numeric value storing the fraction of total variation explained by amplitude variability.

• total_variation: A numeric value storing the amount of total variation.

• n_iterations: An integer value storing the number of iterations performed until convergence;

• call_name: A string storing the name of the function that was used to produce the k-means alignment results;

• call_args: A list containing the exact arguments that were passed to the function call_name that produced this output.

Value

The function as_caps() returns an object of class caps. The function is_caps() returns a boolean which evaluates to TRUE is the input object is of class caps.

Examples

as_caps(sim30_caps)
is_caps(sim30_caps)

fdacluster documentation built on July 9, 2023, 6:45 p.m.