Move each geometry by a small random offset

Shifts each geometry of x as a whole by a random distance along x and along y, each drawn uniformly between -amount and amount. This matches sf::st_jitter() in everything but the random numbers.

Usage

ga_jitter(
  x,
  amount = NULL,
  factor = 0.002,
  seed = sample.int(.Machine$integer.max, 1L)
)

Arguments

x

a GeoArrow geometry array

amount

the most to move a geometry along each axis; length 1 or the same length as x, or NULL to take it from factor

factor

the fraction of the diagonal of the box around x to move a geometry by at most when you give no amount

seed

a whole number that seeds the offsets

Value

a GeoArrow array of the same geometry type and length as x

Details

Without amount, ga_jitter() uses factor times the length of the diagonal of the box around all of x, as sf does. ga_jitter() treats coordinates as planar.

ga_jitter() draws the offsets from its own generator, seeded by seed, two for every row in order. The default seed comes from R's generator, so set.seed() makes the result repeatable, but ga_jitter() never draws the numbers sf would.

ga_jitter() keeps the geometry type of x. It returns null for a null geometry, and for an amount that is NA, negative, or infinite.

See also

Other affine: ga_affine_transform(), ga_rotate_around_center(), ga_rotate_around_centroid(), ga_scale_xy(), ga_skew(), ga_skew_xy(), ga_translate()

Examples

pts <- ga_point(c(0, 10, 10), c(0, 0, 10))

# each point moves by at most 0.5 along each axis
moved <- ga_jitter(pts, amount = 0.5, seed = 1)
as.vector(ga_x(moved)) - c(0, 10, 10)
#> [1]  0.3398737 -0.3860983  0.3883953

# the same seed moves each point the same way
identical(
  as.vector(ga_x(ga_jitter(pts, seed = 42))),
  as.vector(ga_x(ga_jitter(pts, seed = 42)))
)
#> [1] TRUE