Func.motif.multivariate: A function to implment the multivariate motif discovery

In TSMining: Mining Univariate and Multivariate Motifs in Time-Series Data

Description

This function implements the multivariate motif discovery method proposed in A. Vahdatpour, N. Amini, M. Sarrafzadeh. Towards unsupervised activity discovery using multi-dimensional motif detection in time series. IJCAI 2009 21st International Joint Conference on Artificial Intelligence.

Usage

Func.motif.multivariate(motif.list, window.sizes, alpha)

Arguments

motif.list	is a list of lists, each contains the univariate motifs discovered in a univariate time series. The component of motif.list is the results of Func.motif()$Indices, which store the starting position of subsequences of each univariate motif
window.sizes	is a vector containing the length of motifs in each univariate time series. It should have the same order as components in motif.list.
alpha	is a numeric ranging from 0 to 1. It specifies the minimum correlation between two univariate motifs before considered as a multivariate motifs

Value

The function returns a list containing two elements. The first element is Motif, which is a list containing the univarate motif IDs for different multivariate motifs. e.g., if there are two univariate time series and each has 3 motifs, then univariate ID is from 1 to 6. The second element is Info, which is a list storing the information of univariate motifs for different multivariate motifs

Examples

data(test)
#Perform univariate motif discovery for each dimension in the example data
res.1 <- Func.motif(ts = test$TS1, global.norm = TRUE, local.norm = FALSE,
window.size = 10, overlap = 0, w = 5, a = 3, mask.size = 3, eps = .01)
res.2 <- Func.motif(ts = test$TS2, global.norm = TRUE, local.norm = FALSE,
window.size = 20, overlap = 0, w = 5, a = 3, mask.size = 3, eps = .01)
#Perform multivaraite motif discovery
res.multi <- Func.motif.multivariate(motif.list = list(res.1$Indices, res.2$Indices),
window.sizes = c(10,20), alpha = .8)

Example output

================================================================================

TSMining documentation built on May 2, 2019, 3:54 p.m.