geoRts: Geo-reconstruction of Time Series

In InstitutoInvestigacionesEconomicasPUCE/geortsBeta: Reconstruction of Geographical Time Series

Description

This function aproximate by interpolation the values of a time series (unknonw), based in the values of near (geographically) time series.

Usage

geoRts(TS, positions.TS, weights.TS = NULL, positions.RTS = NULL,
  weights.RTS = NULL, D = NULL, ...)

Arguments

TS	is a object of class "ts" (time series), with ncol()>2
positions.TS	is a data frame that containing longitude (lon) and latitude (lat) asociated to time series in TS.
weights.TS	is a data frame that containing weights (of influence) of time series in TS.
positions.RTS	is a data frame that containing values of longitude (lon) and latitude (lat) of each time serie to be reconstructed. If positions.RTS=NULL, then positions.RTS is calculated as a grid of points into a square that contains all points of positions.TS.
weights.RTS	is a data frame that containing weights (of influence) of time series in RTS.
D	is a distance (or disimilarity) matrix between Time Series. If positions.TS is defined and D=NULL then D is calculated as the euclidean distance matrix between positions (i.e. between rows of positions.TS). If positions.TS=NULL, D must be defined, and then positions.TS is calculated (with not a geographical representation) using Multidimensional Scaling (MDS).
...	Extra arguments passed to rts_hull_grid function

Value

returns an object of class "ts", containing the reconstructed time series associate to positions.RTS

InstitutoInvestigacionesEconomicasPUCE/geortsBeta documentation built on June 17, 2020, 7:06 a.m.