inferLID: Infer if dispersion is significant

In andresgmejiar/lbmech: A Package to Study the Mechanics of Landscape and Behavior

Infer if dispersion is significant

Description

Infer whether there exists more or less within- and between-group local and global inequality than would be expected versus if for all observations the values of all other observations were permuted. This tests if local values are significantly above or below what is expected given the global dataset, and if global values are significantly above or below what is expected given an otherwise random distribution.

Usage

inferLID(
  lid,
  w,
  ntrials = 999,
  alpha = 0.05,
  standard = NULL,
  expect = NULL,
  multicore = FALSE,
  ncores = parallel::detectCores() - 1,
  var.stand = FALSE,
  var.exp = FALSE,
  ng.invert = TRUE,
  max.cross = .Machine$integer.max,
  pb = TRUE,
  clear.mem = FALSE
)

Arguments

lid	The list output from the LID function.
w	The same spatial weights matrix used in calculating the lid input.
ntrials	The number of permutations to perform. Default is 999.
alpha	Threshold for significance. Default is alpha = 0.05.
standard	The standards matrix with dimensions length(x) x length(x) used when calculating lid. Ignored if none had been originally provided, otherwise required.
expect	The expectations matrix with dimensions length(x) x length(x) used when calculating lid. Ignored if none had been originally provided, otherwise required.
multicore	Should multithreading be used? Default is FALSE.
ncores	If multicore = TRUE, how many cores should be used? Default is ncores = parallel::detectCores() - 1.
var.stand	Logical. Should the standards be permuted if a matrix was provided? Default is FALSE.
var.exp	Logical. Should the expectations be permuted if a matrix was provided? Default is FALSE.
ng.invert	Does a higher non-group value imply higher between group inequality? Default is TRUE. This is ignored if matrixes were not originally provided, as it is automatically performed.
max.cross	When processing, what is the maximum number of rows that an internal data.table can have? This is generally not a concern unless the number of observations approaches sqrt(.Machine$integer.max)–usually about 2^31 for most systems. Lower values result in a greater number of chunks thus allowing larger data.sets to be calculated.
pb	Logical. Should a progress bar be displayed? Default is pb = FALSE, although if a large dataset is processed that requires adjusting max.cross this can be useful. Ignored if multicore = TRUE.
clear.mem	Logical. Should gc be run in the middle of the calculation? Default is clear.mem but set as TRUE if memory limits are a concern.

Details

The output list can be passed to scatterLID to plot the group and non-group components of local inequality based on the significance classes.

Value

A list with the following entries:

(1) $local A data.table with one column, indicating whether an observation is falls in one of nine categories: Out-group High, Average, or Low for between-group inequality, and In-group High, Average, or Low for within-group inequality based on the significance according to the delta-J statistic in the $stats data.table.

(2) $global A list with four entries, $J_G for the group component of the global inequality, $J_NG for the nongroup, $J for the total, and $Class, containing the significance class for the global dataset. Each of the first three entries themselves contain three entriesL $delta, representing the delta-J statistic, $p, representing its p-value, and $Class, containing the group/non-group class

(3) $stats A data.table containing the number of permutations a randomly-calculated $J_Gi, $J_NGi, or $J_i was above or below the real value

Examples

# Generate dummy observations
x <- runif(10, 1, 100)

# Get distance matrix
dists <- dist(x)

# Get fuzzy weights considering 5 nearest neighbors based on 
# inverse square distance
weights <- makeWeights(dists, bw = 5, 
                       mode = 'adaptive', weighting = 'distance',
                       FUN = function(x) 1/x^2, minval = 0.1,
                       row.stand = 'fuzzy')
                       
# Obtain the 'local gini' value
lid <- LID(x, w = weights, index = 'gini', type = 'local')

# Infer whether values are significant relative to the spatial distribution
# of the neighbots
inference <- inferLID(lid, w = weights, ntrials = 100)

andresgmejiar/lbmech documentation built on Feb. 2, 2025, 12:37 a.m.