Diagnostic Plots for fitted seasonal ARIMA models

Produce diagnostics for fitted seasonal ARIMA models. The method offers several portmanteau tests (including Ljung-Box, Li-McLeod and Box-Pierce), plots of autocorrelations and partial autocorrelations of the residuals, ability to control which graphs are produced (including interactively), as well as their layout.

Usage

# S3 method for Sarima
tsdiag(object, gof.lag = NULL, ask = FALSE, ..., plot = 1:3, layout = NULL)

# if 'object' is produced by stats::arima(), forecast::auto.arima() and
# similar, use the full name, 'tsdiag.Sarima()', in the call. The
# arguments are the same.

Arguments

object

fitted (seasonal) ARIMA model. currently the output of sarima, stats::arima or compatible (e.g., forecast::Arima and forecast::auto.arima). If object is not from sarima, use the full name, tsdiag.Sarima, of the method when calling it, see the examples.

gof.lag

maximal lag for portmanteau tests.

ask

if TRUE present a menu of available plots, see section ‘Details’.

...

not used.

plot

if TRUE all available plots; a vector of positive integers specifies a subset of the available plots.

layout

a list with arguments for graphics::layout for the plots. The default is as for the ARIMA method of stats::tsdiag.

Details

Compute and graph diagnostics for seasonal ARIMA models. For objects of class "Sarima" (produced by sarima) just call the generic, tsdiag. The method can be called also directly on the output from base R's arima() with tsdiag.Sarima() or sarima::tsdiag.Sarima().

The method offers several portmanteau tests (including Ljung-Box, Li-McLeod and Box-Pierce), plots of autocorrelations and partial autocorrelations of the residuals, ability to control which graphs are produced (including interactively), as well as their layout.

The method always makes a correction of the degrees of freedom of the portmanteau tests (roughly, subtracting the number of estimated ARMA parameters). Note that stats::tsdiag doesn't do that.

plot can be TRUE to ask for all plots or a vector of positive integers specifying which plots to consider. Currently the following options are available:

1	residuals
2	acf of residuals
3	p values for Ljung-Box statistic
4	p values for Li-McLeod statistic
5	p values for Box-Pierce statistic
6	pacf of residuals

The default is plot = 1:3, which produces a plot similar to the one from stats::tsdiag (but with adjusted d.f., see above). If plot is TRUE, you probably need also ask = TRUE.

If argument plot is of length two the graphics window is split into 2 equal subwindows. Argument layout can still be used to change this. If argument plot is of length one the graphics window is not split at all.

In interactive sessions, if the number of requested graphs (as specified by argument plot) is larger than the number of graphs specified by the layout (by default 3), the function makes the first graph and then presents a menu of the requested plots.

Argument layout can be used to change the layout of the plot, for example to put two graphs per plot, see the examples. Currently it should be a list of arguments for layout, see ?layout. Don't call layout youself, as that will change the graphics device prematurely.

The computed results are returned (invisibly). This is another difference from stats::tsdiag which doesn't return them.

Value

a list with components:

residuals: residuals
LjungBox: Ljung box test
LiMcLeod: LiMcLeod test
BoxPierce: BoxPierce test

Only components that are actually computed are included, the rest are NULL or absent.

Author

Georgi N. boshnakov

Examples

ap.baseA <- sarima(log(AirPassengers) ~ 
    0 | ma(1, c(-0.3)) + sma(12,1, c(-0.1)) + i(1) + si(12,1), 
    ss.method = "base")
tsdiag(ap.baseA)


## apply the method on objects from arima()
ap.arima <- arima(log(AirPassengers), order = c(0,1,1), seasonal = c(0,1,1))
tsdiag.Sarima(ap.arima)

## use Li-McLeod test instead of Ljung-Box
tsdiag.Sarima(ap.arima, plot = c(1:2,4))

## call R's tsdiag method, for comparison:
tsdiag(ap.arima, plot = c(1:2,4))


## plot only acf and p-values
tsd <- tsdiag.Sarima(ap.arima, plot = c(2:3), layout = list(matrix(1:2, nrow = 2)))

## the results can be used for further calculations:
head(tsd$LjungBox$test, 4)
#>           ChiSq DF    pvalue
#> [1,] 0.03065128  1 0.8610207
#> [2,] 0.11260997  1 0.7371921
#> [3,] 2.45802081  1 0.1169260
#> [4,] 4.30491459  2 0.1161983

if(FALSE) # otherwise R check warns about help page running time > 5s
## plot resid, acf, and p-values, leaving half the space for residuals
tsdiag.Sarima(ap.arima, plot = c(1:3), layout = list(matrix(1:3, nrow = 3),
                                                     heights = c(1,2,2)))

if(FALSE)
## four plots arranged as a 2x2 matrix.
tsdiag.Sarima(ap.arima, plot = c(2:5), layout = list(matrix(1:4, nrow = 2)))