clust.factor: Dynamic Factors Extraction based on Time Series Clustering

In google-trends-v1/gtm: Functions for Economic Nowcasting with Google Trends Data

Description

This function extracts dynamic factors based on time series clustering. Time series are first clustered based on partitional, hierarchical, tadpole or fuzzy methods, and one dynamic factor is then extracted within each cluster, based on either a two-step method as proposed by Doz, Gianone & Reichlin (2011) or a simple aggregation method.

Usage

clust.factor(data=NULL, fac.num=3, method="two-step", clustor.type="partitional", ...)

Arguments

data	a multivariate time series or a numeric matrix to which the clustering and extraction are applied.
fac.num	number of desired clusters, which is also the number of desired factors.
method	method to be applied for extraction of dynamic factors. Options include "two-step" and "aggregation". Defualt is "two-step".
clustyor.type	the type of clustering method to be applied. Options include "partitional", "hierarchical", "tadpole", and "fuzzy". Default is "partitional".
...	extra arguments which are passed to the tsclust function.

Author(s)

Zehua Wu

google-trends-v1/gtm documentation built on June 5, 2019, 5:13 p.m.