cutoff2polygon {SpatialRDD}

Create (treated) polygon from line

Description

Creates an approximation of a "treated/untreated polygon" to assign the status again to each observation after the border has been shifted. The function extends both ends of the provided cutoff to the edge of the (imaginary) bounding box of the provided data (this ensures all observations will be included). Key is that you provide a 2-tuple that indicates in which side of the bounding box each end should go (1st element is the one with lower x-coordinate, i.e. leftern most). Always check the output manually by plotting the polygon (e.g. with tm_shape(your.polygon) + tm_polygons()). If the output polygon looks odd, a first check should be to just switch the elements from the orientation vector around! See vignette(shifting_borders) for details and illustrative examples.

Usage

cutoff2polygon(data, cutoff, orientation = NA, endpoints = c(0, 0))

Arguments

data

study dataset to determine the bounding box (so that all observations are covered by the new polygons) in sf format

cutoff

sf object of the (placebo) cut-off

orientation

in which side of the bounding box does each of the extensions of the cutoff go into? First element refers to endpoint of border with smaller x-coordinate ("westernmost") (takes two of "north", "east", "south", "west" in a vector, e.g. c("west", "north"))

endpoints

at what position on the edge should each polygon end? (vector with two numbers between 0 and 1, where 0.5 e.g. means right in the middle of the respective edge)

Value

a polygon as an sf object

Examples

points_samp.sf <- sf::st_sample(polygon_full, 100) # create points
# make it an sf object bc st_sample only created the geometry list-column (sfc):
points_samp.sf <- sf::st_sf(points_samp.sf)
# add a unique ID to each observation:
points_samp.sf$id <- 1:nrow(points_samp.sf)
cutoff2polygon(data = points_samp.sf, cutoff = cut_off,
orientation = c("west", "west"), endpoints = c(.8, .2))