[R] Help finding the mode and maximum for a specified 'window' of time series data
Moshe Olshansky
m_olshansky at yahoo.com
Mon Jun 16 06:22:41 CEST 2008
Another possibility is to create a matrix
m <- matrix(sample[,2],nrow=4)
and then use apply to compute maximum and mode for every column.
--- On Mon, 16/6/08, jim holtman <jholtman at gmail.com> wrote:
> From: jim holtman <jholtman at gmail.com>
> Subject: Re: [R] Help finding the mode and maximum for a specified 'window' of time series data
> To: "Staples, Angela Dawn" <adstaple at indiana.edu>
> Cc: "r-help at R-project.org" <r-help at r-project.org>
> Received: Monday, 16 June, 2008, 4:05 AM
> This will give you an idea of how you might want to approach
> the problem:
>
> > sample
> <-cbind((1:20)*.25,c(1,1,1,2,2,3,3,3,4,5,6,4,4,3,3,3,2,1,1,1))
> > sample
> [,1] [,2]
> [1,] 0.25 1
> [2,] 0.50 1
> [3,] 0.75 1
> [4,] 1.00 2
> [5,] 1.25 2
> [6,] 1.50 3
> [7,] 1.75 3
> [8,] 2.00 3
> [9,] 2.25 4
> [10,] 2.50 5
> [11,] 2.75 6
> [12,] 3.00 4
> [13,] 3.25 4
> [14,] 3.50 3
> [15,] 3.75 3
> [16,] 4.00 3
> [17,] 4.25 2
> [18,] 4.50 1
> [19,] 4.75 1
> [20,] 5.00 1
> > # range of timing
> > s.r <- range(sample[,1])
> > # create groupings for 1 minute intervals
> > s.cut <- seq(from=floor(s.r[1]),
> to=ceiling(s.r[2]), by=1)
> > s.cut
> [1] 0 1 2 3 4 5
> > # split the data
> > s.split <- split(sample[,2], cut(sample[,1],
> s.cut))
> > s.split
> $`(0,1]`
> [1] 1 1 1 2
>
> $`(1,2]`
> [1] 2 3 3 3
>
> $`(2,3]`
> [1] 4 5 6 4
>
> $`(3,4]`
> [1] 4 3 3 3
>
> $`(4,5]`
> [1] 2 1 1 1
>
> > # determine maximum in interval
> > sapply(s.split, max, na.rm=TRUE)
> (0,1] (1,2] (2,3] (3,4] (4,5]
> 2 3 6 4 2
> > # mode (maximum # of occurances)
> > sapply(s.split, function(x) {
> + .tab <- table(x)
> + as.numeric(names(.tab)[which.max(.tab)])
> + })
> (0,1] (1,2] (2,3] (3,4] (4,5]
> 1 3 4 3 1
> >
> >
>
>
> On Sun, Jun 15, 2008 at 12:10 PM, Staples, Angela Dawn
> <adstaple at indiana.edu> wrote:
> > I am relatively new to R, so apologize up front for my
> long question,
> > particularly if there is too much or too little
> information.
> >
> > I have a large time series data set where each
> subject's behavior was
> > originally coded on .25s intervals for 3min task. I am
> trying to determine
> > if the findings are different depending on the coding
> interval (i.e. Compare
> > .25s to 1s to 5s to 10s).
> >
> > I also need to see if it makes a difference if I use
> the maximal value
> > within the interval or the modal value within the
> interval. In other words,
> > I do not need the average value.
> >
> > Here a sample of my data with time as the first
> column:
> >
> > sample
> <-cbind((1:20)*.25,c(1,1,1,2,2,3,3,3,4,5,6,4,4,3,3,3,2,1,1,1))
> >
> > [,1] [,2]
> > [1,] 0.25 1
> > [2,] 0.50 1
> > [3,] 0.75 1
> > [4,] 1.00 2
> > [5,] 1.25 2
> > [6,] 1.50 3
> > [7,] 1.75 3
> > [8,] 2.00 3
> > [9,] 2.25 4
> > [10,] 2.50 5
> > [11,] 2.75 6
> > [12,] 3.00 4
> > [13,] 3.25 4
> > [14,] 3.50 3
> > [15,] 3.75 3
> > [16,] 4.00 3
> > [17,] 4.25 2
> > [18,] 4.50 1
> > [19,] 4.75 1
> > [20,] 5.00 1
> >
> > I need help returning the maximum and minimum of a
> specified "window" such
> > that the function/loop/etc would return:
> >
> > maximum (window=4, or 1s) would be (2, 3, 6, 4, 2)
> > mode (window=4, or 1s) would be (1, 3, 4, 3, 2)
> >
> > Given the coding conventions in this research area, I
> need the values to be
> > from adjacent, as opposed to overlapping, windows.
> There are likely to be
> > situations where there is no clear mode (1,1,2,2). In
> those cases it would
> > be fine to have 1.5 or 2 returned, but not NA. The
> data file is in the long
> > format with each subject having 720 rows of data.
> I've tried playing with
> > the row indices, but I cannot figure out how to
> 'move' the window.
> >
> > I would appreciate any help/suggestions. Since I'm
> new to this, the
> > suggestion doesn't have to be pretty, I just need
> it to work.
> >
> > Sincerely,
> >
> > Angela
> >
> > ~~~~~~~~~~~~~~~~~~~~~~
> > Angela Staples
> > Doctoral Candidate
> > Psychological and Brain Sciences
> > Indiana University
> > 1101 E. 10th St.
> > Bloomington, IN 47405
> > http://www.indiana.edu/~batessdl/
> >
> > The plural of anecdote is not data.
> > ~ Roger Brinner
> >
> > ______________________________________________
> > R-help at r-project.org mailing list
> > https://stat.ethz.ch/mailman/listinfo/r-help
> > PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> > and provide commented, minimal, self-contained,
> reproducible code.
> >
>
>
>
> --
> Jim Holtman
> Cincinnati, OH
> +1 513 646 9390
>
> What is the problem you are trying to solve?
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained,
> reproducible code.
More information about the R-help
mailing list