Apply Rolling Functions using Circular Statistics — rolling_test

A generic function for applying a function to rolling margins of an array along an additional value.

Usage

distroll_circstats(
  x,
  distance,
  FUN,
  width = NULL,
  min_n = 2,
  align = c("right", "center", "left"),
  w = NULL,
  sort = TRUE,
  ...
)

distroll_confidence(
  x,
  distance,
  w = NULL,
  width = NULL,
  min_n = 2,
  align = c("right", "center", "left"),
  sort = TRUE,
  ...
)

distroll_dispersion(
  x,
  y,
  w = NULL,
  w.y = NULL,
  distance,
  width = NULL,
  min_n = 2,
  align = c("right", "center", "left"),
  sort = TRUE,
  ...
)

distroll_dispersion_sde(
  x,
  y,
  w = NULL,
  w.y = NULL,
  distance,
  width = NULL,
  min_n = 2,
  align = c("right", "center", "left"),
  sort = TRUE,
  ...
)

Arguments

x, y: vectors of numeric values in degrees. length(y) is either 1 or length(x)
distance: numeric. the independent variable along the values in x are sorted, e.g. the plate boundary distances
FUN: the function to be applied
width: numeric. the range across distance on which FUN should be applied on x. If NULL, then width is a number that separates the distances in 10 equal groups.
min_n: integer. The minimum values that should be considered in FUN (2 by default), otherwise NA.
align: specifies whether the index of the result should be left- or right-aligned or centered (default) compared to the rolling window of observations. This argument is only used if width represents widths.
w: numeric. the weighting for x
sort: logical. Should the values be sorted after distance prior to applying the function (TRUE by default).
...: optional arguments to FUN
w.y: numeric. the weighting for y

Value

two-column vectors of (sorted) x and the rolled statistics along distance.

Examples

data("plates")
plate_boundary <- subset(plates, plates$pair == "na-pa")
data("san_andreas")
PoR <- subset(nuvel1, nuvel1$plate.rot == "na")
san_andreas$distance <- distance_from_pb(
  x = san_andreas,
  PoR = PoR,
  pb = plate_boundary,
  tangential = TRUE
)
dat <- san_andreas |> cbind(PoR_shmax(san_andreas, PoR, "right"))

distroll_circstats(dat$azi.PoR, distance = dat$distance, w = 1 / dat$unc, FUN = circular_mean)
#>         distance         x   n
#>  [1,] -9.2265048  68.11323   9
#>  [2,] -7.6044815 127.98953  19
#>  [3,] -5.9824582 155.84338  14
#>  [4,] -4.3604350 154.02577  39
#>  [5,] -2.7384117 138.55072 125
#>  [6,] -1.1163884 138.29357 759
#>  [7,]  0.5056349 142.46616 153
#>  [8,]  2.1276582 147.30864   5
#>  [9,]  3.7496815 178.57158   2
#> [10,]  5.3717047        NA   1
distroll_confidence(dat$azi.PoR, distance = dat$distance, w = 1 / dat$unc)
#> Warning: NaNs produced
#>         distance          x   n
#>  [1,] -9.2265048 147.581596   9
#>  [2,] -7.6044815  44.732416  19
#>  [3,] -5.9824582 101.990075  14
#>  [4,] -4.3604350  43.031293  39
#>  [5,] -2.7384117   9.675720 125
#>  [6,] -1.1163884   4.369003 759
#>  [7,]  0.5056349  11.713927 153
#>  [8,]  2.1276582  20.165812   5
#>  [9,]  3.7496815        NaN   2
#> [10,]  5.3717047         NA   1
distroll_dispersion(dat$azi.PoR, y = 135, distance = dat$distance, w = 1 / dat$unc)
#>         distance          x   n
#>  [1,] -9.2265048 0.56974161   9
#>  [2,] -7.6044815 0.31555788  19
#>  [3,] -5.9824582 0.39396588  14
#>  [4,] -4.3604350 0.32627346  39
#>  [5,] -2.7384117 0.06582746 125
#>  [6,] -1.1163884 0.09488957 759
#>  [7,]  0.5056349 0.11027106 153
#>  [8,]  2.1276582 0.06141144   5
#>  [9,]  3.7496815 0.47800478   2
#> [10,]  5.3717047         NA   1
distroll_dispersion_sde(dat$azi.PoR, y = 135, distance = dat$distance, w = 1 / dat$unc, R = 100)
#>         distance          x   n
#>  [1,] -9.2265048 0.19223776   9
#>  [2,] -7.6044815 0.15681826  19
#>  [3,] -5.9824582 0.21465546  14
#>  [4,] -4.3604350 0.11958988  39
#>  [5,] -2.7384117 0.02090297 125
#>  [6,] -1.1163884 0.01176760 759
#>  [7,]  0.5056349 0.02242678 153
#>  [8,]  2.1276582 0.03415107   5
#>  [9,]  3.7496815 0.05898341   2
#> [10,]  5.3717047         NA   1