display seasonal time series data - Maple Help

Home : Support : Online Help : Statistics : Time Series Analysis Package : TimeSeriesAnalysis/SeasonalSubseriesPlot

TimeSeriesAnalysis[SeasonalSubseriesPlot] - display seasonal time series data

 Calling Sequence SeasonalSubseriesPlot(ts, opts)

Parameters

 ts1 - TimeSeries object opts - (optional) equations of the form optionname = value, where optionname is one of means, starting_period, space, or seasonnames; or a general plot option

Description

 • The SeasonalSubseriesPlot command is used for displaying a plot of the seasonal subseries of a time series; that is, all values for a particular season are shown together. For example, for a monthly time series, this plot shows the January data points of all years together, the February data points, and so on. The seasons are shown next to each other.
 • Optionally, the plot also shows a horizontal line through each subseries, at the mean value for that season.
 • Typically, this command is used for time series having only a single data set, but it is possible to use it for time series having multiple data sets. In that case, the values of all data sets for a particular season are shown together.

Examples

 > $\mathrm{with}\left(\mathrm{TimeSeriesAnalysis}\right):$

Consider the following time series, which exhibits seasonal behavior with period 3.

 > $\mathrm{ts1}:=\mathrm{TimeSeries}\left(\left[4,8,9,3,8,8,5,9,10,7\right],\mathrm{period}=3\right)$
 ${\mathrm{ts1}}{:=}\left[\begin{array}{c}{\mathrm{Time series}}\\ {\mathrm{data set}}\\ {\mathrm{10 rows of data:}}\\ {\mathrm{2005 - 2014}}\end{array}\right]$ (1)
 > $\mathrm{TimeSeriesPlot}\left(\mathrm{ts1}\right)$

 > $\mathrm{SeasonalSubseriesPlot}\left(\mathrm{ts1}\right)$

You can color the lines of the means blue, and make the line charts for the time series thicker. You can also add a legend for the horizontal lines.

 > $\mathrm{SeasonalSubseriesPlot}\left(\mathrm{ts1},\mathrm{thickness}=3,\mathrm{means}=\left[\mathrm{color}="Niagara Blue",\mathrm{legend}="seasonal mean"\right]\right)$

Consider the following monthly time series.

 > $\mathrm{ts2}:=\mathrm{TimeSeries}\left(\left[7,23,21,19,13,46,42,30,31,26,19,9,16,26,17,33,31,46,42,35,45,30,11,17,23,20,15,36,31,55,49,39,36,28,12,11,21,23,27,33,36,49,42,37,33,45,12,7,23,32,25,42,27,52,50,34,41,40,16,14\right],\mathrm{frequency}=\mathrm{monthly},\mathrm{startdate}="2005-09"\right)$
 ${\mathrm{ts2}}{:=}\left[\begin{array}{c}{\mathrm{Time series}}\\ {\mathrm{data set}}\\ {\mathrm{60 rows of data:}}\\ {\mathrm{2005-09-01 - 2010-08-01}}\end{array}\right]$ (2)
 > $\mathrm{TimeSeriesPlot}\left(\mathrm{ts2}\right)$

 > $\mathrm{SeasonalSubseriesPlot}\left(\mathrm{ts2}\right)$

The time series starts in September; it makes sense to start the display with January. You can also specify the names of the months; then you need to make the plot a little wider so that all the names fit. And finally, you increase the space between the months in the plot.

 > $\mathrm{SeasonalSubseriesPlot}\left(\mathrm{ts2},\mathrm{startingperiod}=9,\mathrm{seasonnames}=\left["January","February","March","April","May","June","July","August","September","October","November","December"\right],\mathrm{space}=0.25,\mathrm{size}=\left[800,400\right]\right)$
 >