Re: [R] cumulative time durations of specified periods (chron)

From: Sebastian Luque <spluque_at_gmail.com>
Date: Wed 14 Jun 2006 - 13:16:38 EST

[sorry for my previous empty follow-up -- my fingers got messed up.]

Spencer Graves <spencer.graves@pdf.com> wrote:

> Did you try the following:
>
>> xto-xfrom
> Time in days:
> [1] 1.50 1.75 2.00 2.25

I wanted to find the total amount of time between a series of date/time's corresponding to particular times of the day, not how much time there is between the two series. I ended my coding it as:

"subTime" <- function(from, to, t0, t1)
{

    stopifnot(length(from) == length(to), length(t0) == length(t1))

    startFrom <- chron(sapply(t0, "+", as.numeric(dates(from))))
    endFrom <- chron(sapply(t1, "+", as.numeric(dates(from))))
    startTo <- chron(sapply(t0, "+", as.numeric(dates(to))))
    endTo <- chron(sapply(t1, "+", as.numeric(dates(to))))     fromSeq <- mapply(seq, startFrom, startTo)     toSeq <- mapply(seq, endFrom, endTo)     diffs <- mapply(function(x0, x1, y0, y1) {
        bad <- x1 < y0 | x0 > y1
        full <- x0 >= y0 & x1 <= y1
        fullDiff <- sum(x1[full] - x0[full], na.rm=TRUE)
        r <- x0 >= y0 & x1 > y1 & !bad
        x1[r] <- y1
        truncR <- sum(pmax(x1[r], x0[r]) - x0[r], na.rm=TRUE)
        l <- x0 < y0 & x1 <= y1 & !bad
        x0[l] <- y0
        truncL <- sum(x1[l] - pmax(x0[l], x1[l]), na.rm=TRUE)
        sum(fullDiff, truncR, truncL)

    }, fromSeq, toSeq, from, to)
    diffDims <- list(phases=paste(from, to, sep="-"),
                     periods=paste(t0, t1, sep="-"))
    matrix(diffs, ncol=length(t0), dimnames=diffDims) }

The logicals seemed to be needed to account for truncated periods. It's working as I needed it to, but it looks too convoluted so if somebody finds an easier way to do this, I'd be happy to learn about it. The function returns a matrix:

R> xfrom <- chron(seq(1.25, 11, 3.25))
R> xto <- chron(as.numeric(xfrom) + seq(1.5, 2.25, 0.25))
R> xt0 <- times(c("04:00:00", "11:00:00"))
R> xt1 <- times(c("10:00:00", "16:00:00"))
R> subTime(xfrom, xto, xt0, xt1)
                                         periods
phases                                    04:00:00-10:00:00 11:00:00-16:00:00
  (01/02/70 06:00:00)-(01/03/70 18:00:00)            0.2500            0.4167
  (01/05/70 12:00:00)-(01/07/70 06:00:00)            0.3333            0.2083
  (01/08/70 18:00:00)-(01/10/70 18:00:00)            0.5000            0.4167
  (01/12/70 00:00:00)-(01/14/70 06:00:00)            0.5833            0.4167



> Also, have you seen Gabor Grothendieck and Thomas Petzoldt. "R help
> desk: Date and time classes in R". R News, 4(1):29-32, June 2004.,
> downloadable from www.r-project.org -> "Documentation: Newsletter" as
> well as the "zoo" vignette (see
> "http://finzi.psych.upenn.edu/R/Rhelp02a/archive/67006.html")?

I haven't read the "zoo" vignette yet, but the R News article is one I come back to often.

Thank you,

-- 
Seb

______________________________________________
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Received on Wed Jun 14 13:21:38 2006

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.1.8, at Wed 14 Jun 2006 - 14:11:26 EST.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-help. Please read the posting guide before posting to the list.