Resampling for spatially dependent data

BlockBootstrapGrid: moving-block bootstrap of a grid from overlapping block by block blocks (Hall 1985; Lahiri 2003). StationaryBootstrap: Politis-Romano stationary bootstrap with geometric block lengths of mean \(1/p\) on the circularly wrapped series. BootstrapBands: pointwise percentile and simultaneous (max-|t|) bands from replicate fields. ToroidalShiftTest: correlation of two maps against all (or random) toroidal shifts of the second. Identical to the Python arm morie.fn.spboot; indices are 1-based here.

Usage

BlockBootstrapGrid(
  grid,
  block,
  nboot = 200L,
  seed = 1,
  statistic = mean,
  alpha = 0.05
)

StationaryBootstrap(
  x,
  p,
  nboot = 200L,
  seed = 1,
  statistic = mean,
  alpha = 0.05
)

BootstrapBands(replicates, estimate = NULL, alpha = 0.05)

ToroidalShiftTest(a, b, exact = TRUE, nshift = 199L, seed = 1)

Arguments

grid

Matrix.

block

Block side.

nboot

Replicates.

seed

Philox seed.

statistic

Function of a vector (default mean).

alpha

Level.

x

Series.

p

Block restart probability.

replicates

Matrix of replicate fields (rows).

estimate

Centre of the bands.

a, b

Maps (matrices).

exact

Use all shifts.

nshift

Random shifts.

Value

List.

References

Lahiri, S. N. (2003). Resampling Methods for Dependent Data. Springer.

Politis, D. N. and Romano, J. P. (1994). The stationary bootstrap. JASA 89, 1303-1313.

Upton, G. J. G. and Fingleton, B. (1985). Spatial Data Analysis by Example, Vol. 1. Wiley.

Examples

ToroidalShiftTest(matrix(1:4, 2, byrow = TRUE), matrix(1:4, 2, byrow = TRUE))$statistic
#> [1] 1