Spectral Plots

Interactive plot of the periodogram or the autogregressive spectrum. Produces a ggplot/plotly object for objects of class persephone.

plotSpectrum(
  x,
  tsType = c("original", "sa", "irregular", "residuals"),
  plotType = c("arSpec", "arSpecBars", "periodogram"),
  maxobs = NULL,
  n.freq = NULL,
  order = 30,
  main = NULL,
  interactive = TRUE,
  ...
)

Arguments

x

an object of class persephone.

tsType

character ("original","sa","irregular","residuals") selecting the preferred type of time series to plot. A numeric value (1:4) corresponding to one of these characters is also accepted.

plotType

character ("arSpec","arSpecBars","periodogram") selecting the preferred type of plot, see Details. A numeric value (1:3) corresponding to one of these characters is also accepted.

maxobs

maximum number of observations of the time series. If NULL, maxobs equals the length of the time series for plotTypes "arSpec" and "periodogram" and maxobs=96 for plotType "arSpecBars".

n.freq

the number of frequencies, i.e. the number of points at which to plot. If NULL, n.freq=301 for plotTypes "arSpec" and "periodogram" and n.freq=61 for plotType "arSpecBars".

order

order of the AR model

main

plot title

interactive

If the return value would be a ggplot object, wrap it in plotly::ggplotly before returning.

...

other plotting parameters to affect the plot. Not currently used.

Value

Returns an object of class ggplot or plotly. The seasonal and trading day frequencies are indicated in blue and red. For monthly data, the seasonal frequencies are 1/12, 2/12, ..., 6/12 and the trading day frequencies are 0.348 and 0.432. For quarterly data, the seasonal frequency is 1/4 and the trading day frequencies are 0.044375, 0.08875, 0.294375, 0.33875 and 0.38125.

Details

The following options are available for the parameter plotType.

• arSpec: The autoregressive spectrum similar to JD+ plots. The default settings for the number of frequencies, i.e. the number of points at which to plot, the order of the AR model to be fitted and the maximum number of observations of the time series are n.freq = 301, order = 30 and maxobs='length of the time series' in this case.

• arSpecBars: The autoregressive spectrum similar to X-13ARIMA-SEATS plots. The default settings for the number of frequencies, the order of the AR model to be fitted and the maximum number of observations are n.freq = 61, order = 30 and maxobs=96 in this case.

• periodogram: The raw periodogram similar to JD+ plots. The default settings for the number of frequencies and the maximum number of observations are n.freq = 301 and maxobs='length of the time series' in this case.

Examples

# Monthly Data Example
data(AirPassengers, package = "datasets")
# Generate a persephone object, in this case an x13Single object
obj <- perX13(AirPassengers, "RSA5c")
obj$run()

# Plot the AR-Spectrum after run
# Autoregressive Spectrum of the original series
# (similar to JD+ plots)
# where n.freq = 301 and order = 30
plotSpectrum(obj)


# Autoregressive Spectrum of the original series
# (similar to X-13ARIMA-SEATS plots)
# where n.freq = 61 and order = 30
plotSpectrum(obj, plotType="arSpecBars")


# Periodogram
plotSpectrum(obj, plotType="periodogram")

# same as
plotSpectrum(obj, plotType=3)


# Quarterly Data Example
data(UKgas, package = "datasets")
# Generate a persephone object, in this case a tramoseats object
obj2 <- perTramo(UKgas, "RSAfull")
obj2$run()
plotSpectrum(obj2, tsType="original")

plotSpectrum(obj2, tsType="sa")

plotSpectrum(obj2, tsType="irregular")

plotSpectrum(obj2, tsType="residuals")