Hi R-devel. I've noticed a couple of quirks in the current time/date axis functions (axis.Date, axis.POSIXct, and the equivalents in lattice). Looking at the code, it seems like a fairly ad-hoc approach, often using pretty() on components of the time. This is not always ideal - for example a one-hour interval gets cut into 10-minute chunks rather than the more natural 15-minute chunks (since pretty() doesn't know about the minutes in an hour, etc). Generally the number of tick marks produced varies a lot, and there are a couple of strange cases: try plot(0:1 ~ as.POSIXct(paste(2002:2003,"-02-01",sep="")))
So, I've written a function prettyDate() that attempts to act like pretty(), but with dates and times. Like pretty, it takes arguments 'n' and 'min.n' which specify the desired and minimum number of ticks, respectively. http://pastie.org/751640 By the way, also listed there is an extension of trunc.POSIXt with extra units "weeks", "months", "years", "decades", "centuries". Following is a test of prettyDate() for axis labelling, drawn for a sequence of different time intervals. source("http://pastie.org/751640.txt";) steps <- list("10 secs", "1 min", "5 mins", "10 mins", "15 mins", "30 mins", "1 hour", "3 hours", "6 hours", "12 hours", "1 DSTday", "1 week", "2 weeks", "1 month", "3 months", "6 months", "1 year", "2 years", "5 years", "10 years", "20 years", "50 years", "100 years") names(steps) <- paste("span =", unlist(steps)) from <- as.POSIXct("2002-02-02 02:02") devAskNewPage(TRUE) lapply(steps, function(s) { times <- seq(from, by = s, length = 2) plot(0:1 ~ times, yaxt = "n", ylab = "") x <- mean(par("usr")[1:2]) text(x, 0.5, paste("span:", s), cex = 2) text(x, 0.33, paste(format(times), collapse="\n")) text(x, 0.05, "current axis.POSIXct") text(x, 0.95, "proposed new prettyDate axis") ## draw new proposed axis function at top of plot timelim <- par("usr")[1:2] mostattributes(timelim) <- attributes(from) axis(side = 3, at = prettyDate(timelim), labels = prettyDate(timelim, do.format=TRUE)) }) devAskNewPage(FALSE) Is it appropriate / desirable for this to be incorporated into R? -- Felix Andrews / 安福立 Postdoctoral Fellow Integrated Catchment Assessment and Management (iCAM) Centre Fenner School of Environment and Society [Bldg 48a] The Australian National University Canberra ACT 0200 Australia M: +61 410 400 963 T: + 61 2 6125 4670 E: felix.andr...@anu.edu.au CRICOS Provider No. 00120C -- http://www.neurofractal.org/felix/ ______________________________________________ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
