[R] question about the tisPlot function in package tis

Jeffrey J. Hallman jhallman at frb.gov
Tue Dec 2 23:19:04 CET 2008


Kyle.Matoba at sf.frb.org writes:

> List,
>
> I am using the 'tisPlot' function in Jeff Hallman's excellent tis package 
> and was hoping that someone could spare me from having to dig into the 
> code of his 'tisPlot' function.  So far as I can tell, the preferred 
> method of controlling the plotting of the x-axis is using the 'xTickFreq' 
> and 'xTickSkip' options.  Unfortunately, the where the data ends up on the 
> resultant plot is not invariant to this choice, as indicated in the 
> documentation of the 'xTickFreq' option.  Has anyone adjusted the 
> frequency of their x-axis label using this package before?  I imagine that 
> there must be a better way of doing what I am trying to accomplish, but if 
> not, does anyone have a suggestion on how to work around this without 
> digging into the function itself?
>
> ## an example
>
> library(tis)
>
> strt <- ti(20000101, "monthly")
> dat  <- tis(runif(9*12 - 1), start=strt) # ends in 11/2008
>
> par(mfrow=c(2,1))
> tisPlot(dat, xTickFreq="monthly", xTickSkip=6) # graph ends in 11/2008, 
> but looks to start some time in late 1999
> tisPlot(dat, xTickFreq="monthly", xTickSkip=12) # graph looks to end 
> around 3/2009

Try this:

tisPlot(dat, xTickFreq = "annual", xMinorTickFreq = "monthly")


I don't understand what you mean by "where the data ends up on the 
> resultant plot is not invariant to this choice, as indicated in the 
> documentation of the 'xTickFreq' option."  The documentation for xTickFreq
> says:

xTickFreq: a string like the ones returned by 'tifName'. This argument
          and 'xTickSkip' jointly specify locations for labeled x axis
          ticks as follow: (i) find the 'ti''s of the given frequency
          that correspond to 'xAxisMin' and 'xAxisMax', then (ii)
          including those as endpoints, generate a sequence of every
          'xTickSkip''th 'ti' between them.

          Two special strings can also be given.  "none" means no
          labelled tick marks, while "auto" tries (not always
          successfully) to come up with reasonable tick locations
          automatically.  "auto" also overrides any 'xTickSkip'
          setting.

          The default is "auto". 

which does not indicate that your data points are moved.  At any rate, please
try my suggestion above, which gives a pretty nice axis for your time range.

Jeff



More information about the R-help mailing list