gof_mc: Polygons Spatial Association Test - Global Envelope
In lcgodoy/tpsa: Testing Polygons Spatial Association

gof_mc

R Documentation

Polygons Spatial Association Test - Global Envelope

Description

A Monte Carlo test to verify if two sets of polygons are associated based in a global envelope of the functions K_{12}(d) and L_{12}(d) using different test statistics.

Usage

gof_mc(
  obj_sp1,
  obj_sp2,
  n_sim = 499L,
  unique_bbox = NULL,
  alpha = 0.01,
  H = "L",
  ts = "SMAD",
  distances = NULL,
  fixed = FALSE,
  method = "hausdorff"
)

Arguments

`obj_sp1`	an object from class `SpatialPolygons` or `SpatialPointsDataFrame`
`obj_sp2`	an object from class `SpatialPolygons` or `SpatialPointsDataFrame`
`n_sim`	an `integer` corresponding to the number of Monte Carlo simulations for the test
`unique_bbox`	a `matrix` `2 \times 2` corresponding to the boundary box that contains both sets
`alpha`	a `numeric` indicating the confidence level.
`H`	a `character` indicating the function to be used. Possible entries are: `'K'` or `'L'`.
`ts`	a `character` associated to a test statistic. Inputs acepted: `c('IM', 'MAD', 'SIM', 'SMAD', 'IMDQ', 'MADDQ')`.
`distances`	a `numeric vector` indicating the distances to evaluate `H(d)`. If `NULL` then the range considered goes from 5 observed inside the `unique_bbox`.
`fixed`	a `boolean` indicating if the first pattern should be fixed on the toroidal shift or the first will be fixed in half of iterations and then the other one. `TRUE` or `FALSE`, respectively.
`method`	a `character` specifying which kind of distance will be used to evalueate `H`. Also, there is an option of using areas. Options available: `c('hausdorff', 'euclidean', 'area')`.

Value

a list from class gof_test, with values:

p_value: a numeric scalar giving the p-value of the test
mc_sample: a numeric vector giving the test statistic for each of the Monte Carlo simulations
mc_funct: a matrix where each line correspond to the function (K or L) estimated for the Monte Carlo simulations
distances: numeric vector containing the distances where mc_func were evaluated.
alpha: a numeric scalar giving the significance level
rejects: a logical scalar, TRUE if the null hypothesis is reject