Skip to contents

Apply a function over a rolling window

Source: R/rolling.R
rolling.Rd

rolling applies a funtion (FUN) over a moving window (see .width and .direction) for each element of x using zoo::rollapply().

Usage

rolling(
  x,
  FUN = mean,
  ...,
  .width = 3,
  .direction = "backward",
  .fill = NA,
  .min_non_na = 0
)

Arguments

x

A vector.

FUN

A function to be applied to each window of x. If is a character string, it will be converted to a function using get(), unless it is in "sum", "mean", "min", or "max" which have built-in vectorized (faster) implementations. Default is "mean".

...

Additional arguments to be passed to FUN.

.width

A numeric value indicating the width of the window. See zoo::rollapply() for more details. Default is 3.

.direction

A character string or numeric value indicating the direction of the window. Options are "backward" (1), "forward" (-1), or "center" (0). See the align argument in zoo::rollapply() for more details. Default is "backward".

.fill

(Optional) A single value to be used for filling in any values that are outside of the window. See zoo::rollapply() for more details. Default is NA.

.min_non_na

(Optional) A single numeric value indicating the minimum number of observations required to compute a value. Default is 0.

Value

A vector of the same length as x and type of the output of FUN.

Details

  • Be sure to fill any gaps in x and ensure x is sorted, otherwise the window may be applied incorrectly.

  • Values outside of the window are replaced with .fill.

  • .min_non_na is passed to do_if_enough() to only apply FUN if enough values are present in the window.

Examples

x <- c(1, 2, 3, 4, 5)
x |> rolling(mean, .width = 2)
#> [1]  NA 1.5 2.5 3.5 4.5
x |> rolling(mean, .width = 2, .direction = "forward", .fill = -1)
#> [1]  1.5  2.5  3.5  4.5 -1.0