Creating_gaps: Creating gaps in multivariate time series

In FSMUMI: Imputation of Time Series Based on Fuzzy Logic

Description

This function builds gaps (with the same size) in multivariate time series. The size of each gap is defined as a percentage of input data length. By default, the position of each gap is randomly chosen.

Usage

Creating_gaps(data, rate, ngaps, begin = NULL)

Arguments

data	input multivariate time series
rate	size of a gap, as a percentage of input time series size
ngaps	number of gaps to create on each signal
begin	the starting position of a gap (random by default)

Value

Creating_gaps returns a multivariate time series with one gap per signal.

Author(s)

Thi-Thu-Hong Phan, Andre Bigand, Emilie Poisson-Caillault

Examples

data(dataFSMUMI)
X <- dataFSMUMI[1:5000,]; #reduction for demo
rate <- 0.1
ngaps <- 1
incompleted_signal <- Creating_gaps(X, rate,1)

FSMUMI documentation built on May 2, 2019, 12:40 p.m.