tsEvaTransformSeriesToStatSeasonal_ciPercentile {RtsEva}

tsEvaTransformSeriesToStatSeasonal_ciPercentile

Description

This function decomposes a time series into a season-dependent trend and a season-dependent standard deviation. The season-dependent amplitude is given by a seasonal factor multiplied by a slowly varying percentile.

Usage

tsEvaTransformSeriesToStatSeasonal_ciPercentile(
  timeStamps,
  series,
  timeWindow,
  percentile
)

Arguments

timeStamps

A vector of time stamps for the time series.

series

The original time series.

timeWindow

The length of the moving window used for trend estimation.

percentile

The percentile value used for computing the slowly varying percentile.

Value

A list containing the following components:

runningStatsMulteplicity

The size of each sample used to compute the average

stationarySeries

The transformed stationary series

trendSeries

The trend series

trendSeriesNonSeasonal

The non-seasonal trend series

stdDevSeries

The season-dependent standard deviation series

stdDevSeriesNonSeasonal

The non-seasonal standard deviation series

trendError

The error on the trend

stdDevError

The error on the standard deviation

statSer3Mom

The 3rd moment of the transformed stationary series

statSer4Mom

The 4th moment of the transformed stationary series

nonStatSeries

The original non-stationary series

Regime

The regime of the trend seasonality

timeStamps

The time stamps

trendNonSeasonalError

The error on the non-seasonal trend

stdDevNonSeasonalError

The error on the non-seasonal standard deviation

trendSeasonalError

The error on the seasonal trend

stdDevSeasonalError

The error on the seasonal standard deviation

this function decomposes the series into a season-dependent trend and a

season-dependent standard deviation. The season-dependent standard deviation is given by a seasonal factor multiplied by a slowly varying standard deviation. transformation non stationary -> stationary x(t) = (y(t) - trend(t) - ssn_trend(t))/(stdDev(t)*ssn_stdDev(t)) transformation stationary -> non stationary y(t) = stdDev(t)*ssn_stdDev(t)*x(t) + trend(t) + ssn_trend(t) trasfData.trendSeries = trend(t) + ssn_trend(t) trasfData.stdDevSeries = stdDev(t)*ssn_stdDev(t)

Examples

timeAndSeries <- ArdecheStMartin
timeStamps <- ArdecheStMartin[,1]
series <- ArdecheStMartin[,2]
#select only the 5 latest years
yrs <- as.integer(format(timeStamps, "%Y"))
tokeep <- which(yrs>=2015)
timeStamps <- timeStamps[tokeep]
series <- series[tokeep]
timeWindow <- 365 # 1 year
percentile <- 90
result <- tsEvaTransformSeriesToStatSeasonal_ciPercentile(timeStamps,
series, timeWindow, percentile)
plot(result$trendSeries)