Message-ID: <BANLkTi=oLf82v-rbOr3-0_d+_1r4iLMU+A@mail.gmail.com>
Date: 2011-06-25T22:07:26Z
From: Gabor Grothendieck
Subject: Moving average in a data table
In-Reply-To: <4E0633DD.2010304@naumenko.ca>
On Sat, Jun 25, 2011 at 3:15 PM, Roman Naumenko <roman at naumenko.ca> wrote:
> Hi,
>
> I'm trying to figure out common approach on calculating MA on a dataset that
> contains column "time".
>
> After digging around, I believe functions rollmean and rollaply should be
> used.
> However I don't quite understand the requirements for the underlying data.
> Should it be zoo object type? Formatted in a special way?
>
> As an example, I'm looking to get calculated avg=MA(variable) over 15 sec
> period on "time_sec" column:
>
> ?date ? ? ? ? ? variable ? time_sec ? ? ? ? avg
> ?2011-05-17 ? ? 132.55 ? ? 11:29:59.579 ? ? 132.55
> ?2011-05-17 ? ? 132.65 ? ? 11:29:59.946 ? ? 132.60
> ?2011-05-17 ? ? 132.5 ? ? ?11:29:59.946 ? ? 132.57
> ?2011-05-17 ? ? 132.5 ? ? ?11:29:59.946 ? ? 132.55
> ?2011-05-17 ? ? 132.55 ? ? 11:29:59.946 ? ? 132.55
> ?2011-05-17 ? ? 132.6 ? ? ?11:29:59.946 ? ? 132.56
> ?2011-05-17 ? ? 132.55 ? ? 11:29:59.946 ? ? 132.56
> ?2011-05-17 ? ? 132.65 ? ? 11:29:59.947 ? ? 132.57
> ?2011-05-17 ? ? 132.85 ? ? 11:30:00.45 ? ? ?132.60
> ?2011-05-17 ? ? 132.9 ? ? ?11:30:00.45 ? ? ?132.63
> ?2011-05-17 ? ? 133.05 ? ? 11:30:00.45 ? ? ?132.67
> ?2011-05-17 ? ? 132.2 ? ? ?11:30:00.45 ? ? ?132.63
> ?2011-05-17 ? ? 132.5 ? ? ?11:30:00.45 ? ? ?132.62
> ?2011-05-17 ? ? 132.7 ? ? ?11:30:00.50 ? ? ?132.63
> ?2011-05-17 ? ? 132.75 ? ? 11:30:00.57 ? ? ?132.63
> ?2011-05-17 ? ? 132.55 ? ? 11:30:00.70 ? ? ?132.63
> ?2011-05-17 ? ? 132.25 ? ? 11:30:00.70 ? ? ?132.61
> ?2011-05-17 ? ? 132.25 ? ? 11:30:00.71 ? ? ?132.59
> ?2011-05-17 ? ? 132.35 ? ? 11:30:00.173 ? ? 132.57
> ?2011-05-17 ? ? 132.45 ? ? 11:30:00.173 ? ? 132.57
>
rollapply and rollmean are for fixed offsets such as 5 rows before and
after. For this problem modify the following depending on your
precise requirements:
Lines <- "date variable time_sec avg
2011-05-17 132.55 11:29:59.579 132.55
2011-05-17 132.65 11:29:59.946 132.60
2011-05-17 132.5 11:29:59.946 132.57
2011-05-17 132.5 11:29:59.946 132.55
2011-05-17 132.55 11:29:59.946 132.55
2011-05-17 132.6 11:29:59.946 132.56
2011-05-17 132.55 11:29:59.946 132.56
2011-05-17 132.65 11:29:59.947 132.57
2011-05-17 132.85 11:30:00.45 132.60
2011-05-17 132.9 11:30:00.45 132.63
2011-05-17 133.05 11:30:00.45 132.67
2011-05-17 132.2 11:30:00.45 132.63
2011-05-17 132.5 11:30:00.45 132.62
2011-05-17 132.7 11:30:00.50 132.63
2011-05-17 132.75 11:30:00.57 132.63
2011-05-17 132.55 11:30:00.70 132.63
2011-05-17 132.25 11:30:00.70 132.61
2011-05-17 132.25 11:30:00.71 132.59
2011-05-17 132.35 11:30:00.173 132.57
2011-05-17 132.45 11:30:00.173 132.57"
DF <- read.table(textConnection(Lines), header = TRUE)
DF <- transform(DF, datetime = as.POSIXct(paste(date, time_sec)))
f <- function(i) {
is.near <- abs(as.numeric(DF$datetime[i] - DF$datetime)) < 7.5
mean(DF$variable[is.near])
}
sapply(1:nrow(DF), f)
--
Statistics & Software Consulting
GKX Group, GKX Associates Inc.
tel: 1-877-GKX-GROUP
email: ggrothendieck at gmail.com