sppt_diff: Performs a Spatial Point Pattern Test following Wheeler et...

In wsteenbeek/sppt: Spatial Point Pattern Test

Description

As opposed to the traditional SPPT, this test calculates the difference in the proportions using either a variety of Chi-square proportions tests or using Fisher's exact test, and then the p-values are adjusted for multiple comparisons. See https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3111822 for a description and example.

Usage

sppt_diff(
  p1.sp,
  p2.sp,
  uoa.sp,
  conf_level = 95,
  test = "Chi2_Nmin1",
  adj = "BY"
)

Arguments

p1.sp	the first points of type SpatialPoints*, order does not matter
p2.sp	the second points of type SpatialPoints*, order does not matter
uoa.sp	the units of analysis of type SpatialPolygons*
conf_level	confidence interval, default = 95
test	test to conduct, choice of "Chi2_Nmin1", "Chi2", "Yates", or "Fisher", default = "Chi2_Nmin1"
adj	method to adjust p-values (as per stats::p.adjust), default "BY". To turn off, set to "none".

Value

Returns uoa.sp (a spatial polygons data frame) including sppt_diff outcomes.

Examples

# Plot areas, base points data, and test points data
plot(areas.sp)
text(coordinates(areas.sp), label = areas.sp$ID)
text(coordinates(points1.sp), label = points1.sp$ID, col="blue")
text(coordinates(points2.sp), label = points2.sp$ID, col="red")

output <- sppt_diff(points1.sp, points2.sp, areas.sp, test="Chi2_Nmin1")
summary_sppt(output)
output@data

output <- sppt_diff(points1.sp, points2.sp, areas.sp, test="Fisher")
summary_sppt(output)
output@data

# sppt_diff(points1.sp, points2.sp, areas.sp, test="????") # should get error message

# Vancouver data
set.seed(9866)
myoutput <- sppt_diff(vancouver_points1.sp, vancouver_points2.sp, vancouver_areas.sp)
summary_sppt(myoutput)

# The sppt_diff() outcomes without correction for multiple comparisons
# are close to sppt_boot() outcomes:
set.seed(9866)
myoutput1 <- sppt_diff(vancouver_points1.sp, vancouver_points2.sp,
             vancouver_areas.sp, test = "Chi2_Nmin1", adj = "none")
summary_sppt(myoutput1)

set.seed(9866)
myoutput2 <- sppt_boot(vancouver_points1.sp, vancouver_points2.sp,
             vancouver_areas.sp)
summary_sppt(myoutput2)

wsteenbeek/sppt documentation built on Oct. 16, 2020, 6:11 p.m.