runFun {TTR}R Documentation

Analysis of Running/Rolling/Moving Windows

Description

Various functions to analyze data over a moving window of periods.

Usage

  runSum(x, n=10)
  runMin(x, n=10)
  runMax(x, n=10)
  runMean(x, n=10)
  runMedian(x, n=10, low=FALSE, high=FALSE)
  runCov(x, y, n=10, use="all.obs", sample=TRUE)
  runCor(x, y, n=10, use="all.obs", sample=TRUE)
  runVar(x, n=10, sample=TRUE)
  runSD(x, n=10, sample=TRUE)
  runMAD(x, n=10, center=runMedian(x, n), stat="median",
         constant=1.4826, low=FALSE, high=FALSE)

Arguments

x Object able to be coerced to a matrix.
y Object able to be coerced to a matrix.
n Number of periods to use in the window.
sample Logical, sample covariance if TRUE (denominator of n-1)
use Only "all.obs" currently implemented.
low Logical, if TRUE, compute the 'low-median'; i.e., take the smaller of the two middle values for even sample sizes.
high Logical, if TRUE, compute the 'high-median'; i.e., take the larger of the two middle values for even sample sizes.
center The values to use as the measure of central tendency, around which to calculate deviations.
stat Statistic to calculate, one of 'median' or 'mean' (e.g. median absolute deviation or mean absolute deviation, respectively.)
constant Scale factor applied to approximate the standard deviation.

Value

runSum
runMin
runMax
runMean
runMedian
runCov
runCor
runVar
runSD
runMAD

returns median/mean absolute deviations over a n-period moving window.

Author(s)

Josh Ulrich


[Package TTR version 0.14-0 Index]