Overplotting: plot() invocation looks ugly ... suggestions?

Try:

matplot(levels(data$Quarter), matrix(data$Consumption, 4), type = "o")


On 7/24/06, John McHenry <john_d_mchenry at yahoo.com> wrote:

Hi WizaRds,

I'd like to overplot UK fuel consumption per quarter over the course of five years.
Sounds simple enough?

Unless I'm missing something, the following seems very involved for what I'm trying to do. Any suggestions on simplifications?

The way I did it is awkward mainly because of the first call to plot ... but isn't this necessary, especially to set limits for the plot?

The second call to plot(), in conjunction with by(), seems to be natural enough, and, IMHO, seems to be readable and succinct.

   data<- read.table(textConnection("Year    Quarter        Consumption
   1965    1        874
   1965    2        679
   1965    3        616
   1965    4        816

   1966    1        866
   1966    2        700
   1966    3        603
   1966    4        814

   1967    1        843
   1967    2        719
   1967    3        594
   1967    4        819

   1968    1        906
   1968    2        703
   1968    3        634
   1968    4        844

   1969    1        952
   1969    2        745
   1969    3        635
   1969    4        871"), header=TRUE)
   data$Quarter<- as.factor(data$Quarter)
   #
   # what follows is only marginally less involved than using a for loop
   # (the culprit is, in part, the need to make the first, type="n", call to plot()):
   windows(width=12,height=6)
   with(data, plot(levels(Quarter), Consumption[Year==Year[1]], ylim=c(min(Consumption), max(Consumption)), type="n"))
   with(data, by(Consumption, Year, function(x) lines(levels(Quarter), x, type="o")))

Thanks,

Jack.



---------------------------------
Groups are talking. We?re listening. Check out the handy changes to Yahoo! Groups.
       [[alternative HTML version deleted]]