Rolling Means/Maximums/Medians/Sums

Generic functions for computing rolling means, maximums, medians, and sums of ordered observations.

rollmean(x, k, fill = if (na.pad) NA, na.pad = FALSE, 
  align = c("center", "left", "right"), ...)

rollmax(x, k, fill = if (na.pad) NA, na.pad = FALSE, 
  align = c("center", "left", "right"), ...)

rollmedian(x, k, fill = if (na.pad) NA, na.pad = FALSE, 
  align = c("center", "left", "right"), ...)

rollsum(x, k, fill = if (na.pad) NA, na.pad = FALSE, 
  align = c("center", "left", "right"), ...)

rollmeanr(..., align = "right")
rollmaxr(..., align = "right")
rollmedianr(..., align = "right")
rollsumr(..., align = "right")

Arguments

x

an object (representing a series of observations).

k

integer width of the rolling window. Must be odd for rollmedian.

fill

a three-component vector or list (recycled otherwise) providing filling values at the left/within/to the right of the data range. See the fill argument of na.fill for details.

na.pad

deprecated. Use fill = NA instead of na.pad = TRUE.

align

character specifying whether the index of the result should be left- or right-aligned or centered (default) compared to the rolling window of observations.

...

Further arguments passed to methods.

Details

These functions compute rolling means, maximums, medians, and sums respectively and are thus similar to rollapply but are optimized for speed.

Currently, there are methods for "zoo" and "ts" series and default methods. The default method of rollmedian is an interface to runmed. The default methods of rollmean and rollsum do not handle inputs that contain NAs. In such cases, use rollapply instead.

If x is of length 0, x is returned unmodified.

Value

An object of the same class as x with the rolling mean/max/median/sum.

See also

rollapply, zoo, na.fill

Examples

suppressWarnings(RNGversion("3.5.0"))
set.seed(1)

x.Date <- as.Date(paste(2004, rep(1:4, 4:1), sample(1:28, 10), sep = "-"))
x <- zoo(rnorm(12), x.Date)

## rolling operations for univariate series
rollmean(x, 3)
#> 2004-01-11 2004-01-15 2004-01-23 2004-02-05 2004-02-21 2004-02-25 2004-03-13 
#>  0.1350951  0.6005117  0.3362392  0.5940580  0.5320787 -0.1043585 -0.8153657 
#> 2004-03-14 
#> -0.5703365 
rollmax(x, 3)
#> 2004-01-11 2004-01-15 2004-01-23 2004-02-05 2004-02-21 2004-02-25 2004-03-13 
#>  0.7383247  0.7383247  0.7383247  1.5117812  1.5117812  1.5117812  0.3898432 
#> 2004-03-14 
#>  1.1249309 
rollmedian(x, 3)
#> 2004-01-11 2004-01-15 2004-01-23 2004-02-05 2004-02-21 2004-02-25 2004-03-13 
#>  0.4874291  0.5757814  0.5757814  0.5757814  0.3898432  0.3898432 -0.6212406 
#> 2004-03-14 
#> -0.6212406 
rollsum(x, 3)
#> 2004-01-11 2004-01-15 2004-01-23 2004-02-05 2004-02-21 2004-02-25 2004-03-13 
#>  0.4052854  1.8015351  1.0087177  1.7821741  1.5962360 -0.3130755 -2.4460972 
#> 2004-03-14 
#> -1.7110095 

## rolling operations for multivariate series
xm <- zoo(matrix(1:12, 4, 3), x.Date[1:4])
rollmean(xm, 3)
#>                  
#> 2004-01-11 2 6 10
#> 2004-01-15 3 7 11
rollmax(xm, 3)
#>                  
#> 2004-01-11 3 7 11
#> 2004-01-15 4 8 12
rollmedian(xm, 3)
#>                  
#> 2004-01-11 2 6 10
#> 2004-01-15 3 7 11
rollsum(xm, 3)
#>                   
#> 2004-01-11 6 18 30
#> 2004-01-15 9 21 33

## rollapply vs. dedicated rollmean
rollapply(xm, 3, mean) # uses rollmean
#>                  
#> 2004-01-11 2 6 10
#> 2004-01-15 3 7 11
rollapply(xm, 3, function(x) mean(x)) # does not use rollmean
#>                  
#> 2004-01-11 2 6 10
#> 2004-01-15 3 7 11