mst.all: Minimum spanning tree for all regions

In jfrench/smerc: Statistical Methods for Regional Counts

Minimum spanning tree for all regions

Description

mst.all finds the set of connected regions that maximize the spatial scan statistic (the likelihood ratio test statistic) from each starting region, subject to relevant constraints. The function can be used to construct candidate zones for the dynamic minimum spanning tree (dmst), early stopping dynamic minimum spanning tree (edmst), double connected spatial scan test (dc), and maximum linkage (mlink) spatial scan test.

Usage

mst.all(
  neighbors,
  cases,
  pop,
  w,
  ex,
  ty,
  max_pop,
  type = "maxonly",
  nlinks = "one",
  early = FALSE,
  cl = NULL,
  progress = FALSE
)

Arguments

neighbors	A list containing the vector of neighbors for each region (in ascending order of distance from the region). The starting region itself is included among the neighbors.
cases	The number of cases observed in each region.
pop	The population size associated with each region.
w	A binary spatial adjacency matrix for the regions.
ex	The expected number of cases for each region. The default is calculated under the constant risk hypothesis.
ty	The total number of cases in the study area.
max_pop	The population upperbound (in total population) for a candidate zone.
type	One of "maxonly", "pruned", or "all". See Details.
nlinks	A character vector. The options are "one", "two", or "max". See Details.
early	A logical value indicating whether the "early" stopping criterion should be used. If TRUE, each sequence is stopped when the next potential zone doesn't produce a test statistic larger than the current zone. The default is FALSE.
cl	A cluster object created by makeCluster, or an integer to indicate number of child-processes (integer values are ignored on Windows) for parallel evaluations (see Details on performance). It can also be "future" to use a future backend (see Details), NULL (default) refers to sequential evaluation.
progress	A logical value indicating whether a progress bar should be displayed. The default is TRUE.

Details

This function is not intended to be used by users directly. Consequently, it prioritizes efficiency over user friendliness.

type is a character vector indicating what should be returned by the function. If type = "maxonly", then the maximum test statistic from each starting region is returned . If type = "pruned", the function returns a list that includes the location ids, test statistic, total cases, expected cases, and total population for the zone with the maximum test statistic for each starting region. If type = "all", the function returns a list of lists that includes the location ids, test statistic, total cases, expected cases, and total population for the sequence of candidate zones associated with each starting region.

If nlinks = "one", then a region only needs to be connected to one other region in the current zone to be considered for inclusion in the next zone. If nlinks = "two", then the region must be connected to at least two other regions in the current zone. If nlinks = "max", then only regions with the maximum number of connections to the current zone are considered for inclusion in the next zone.

Value

Returns a list of relevant information. See Details.

Author(s)

Joshua French

References

Assuncao, R.M., Costa, M.A., Tavares, A. and Neto, S.J.F. (2006). Fast detection of arbitrarily shaped disease clusters, Statistics in Medicine, 25, 723-742. <doi:10.1002/sim.2411>

Costa, M.A. and Assuncao, R.M. and Kulldorff, M. (2012) Constrained spanning tree algorithms for irregularly-shaped spatial clustering, Computational Statistics & Data Analysis, 56(6), 1771-1783. <doi:10.1016/j.csda.2011.11.001>

Examples

# load data
data(nydf)
data(nyw)

# create relevant data
coords <- nydf[, c("longitude", "latitude")]
cases <- floor(nydf$cases)
pop <- nydf$population
w <- nyw
ex <- sum(cases) / sum(pop) * pop
ubpop <- 0.5
ubd <- 0.5
ty <- sum(cases) # total number of cases
# intercentroid distances
d <- gedist(as.matrix(coords), longlat = TRUE)
# upperbound for population in zone
max_pop <- ubpop * sum(pop)
# upperbound for distance between centroids in zone
max_dist <- ubd * max(d)
# create list of neighbors for each region
# (inclusive of region itself)
all_neighbors <- nndist(d, ubd)
# find the dmst max zone
## Not run: 
out <- mst.all(all_neighbors, cases, pop, w, ex, ty, max_pop,
  type = "maxonly"
)
head(out)

out <- mst.all(all_neighbors, cases, pop, w, ex, ty, max_pop,
  type = "pruned"
)
head(out)

## End(Not run)

jfrench/smerc documentation built on Oct. 27, 2024, 5:13 p.m.