In mikejohnson51/AOI: Areas of Interest

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  out.width = "100%",
  dev = "jpeg"
)

library(AOI)
library(dplyr)

Basic Usage

The core functions described here are aoi_ext which builds a specific extent around a given POINT / coordinate set and discritize which discritizes an extent/object.

To start, lets build a 1000 meter extent around the location called "Fort Collins":

aoi_ext(geo = "Fort Collins", wh = 1000)

From this, we can see a extent was created!

Add bbox

As with AOI::geocode, the returned object can be represented as a POLYGON by adding bbox = TRUE. This is helpful for mapping:

aoi_ext(geo = "Fort Collins", wh = 1000, bbox = TRUE) %>% 
  mapview::mapview()

Modify both diminisons

Providing a single value to the wh argument generates a square extent. To find a rectangular extent, a two value vector can be passed:

aoi_ext(geo = "Fort Collins", wh = c(2000, 1000), bbox = TRUE) %>% 
  mapview::mapview()

Use XY

We have relied on providing a place that requires AOI::geocode to be run. Instead, we can provide a direct coordinate set as an xy vector:

aoi_ext(xy = c(x = -119, y = 36), wh = 1000)

Units

By default all measures are made in the default AOI unit:

(AOI:::default_units)

We can override these by defining the units argument in the aoi_ext signiture. Here we see that a 1km and 1000m wh argument produce identical results:

identical( aoi_ext(geo = "Fort Collins", wh = 1, units = "km"), 
           aoi_ext(geo = "Fort Collins", wh = 1000) )

Area discritizations

In many applications, getting the extent is only part of the challenge. The other is defining a discretizion of a region. This can be done with the AOI::discritize function:

aoi_ext("Fort Collins", wh = c(10000, 20000)) %>% 
  discritize()

Diminsions

By default, the function uses the AOI default discritization:

(AOI:::default_dim)

This can be changed using the dim argument:

aoi_ext("Fort Collins", wh = c(10000, 20000)) %>% 
  discritize(dim = 1024)

CRS

The default expected input CRS is EPSG:4326. This is also the default output. Both can be changed depending on the desired input and output.

• If a CRS can be extracted from the input, then it is used.
• If a CRS cannot be extracted from the input, then EPSG:4326 is assumed
• The assumed CRS can be overridden with the in_crs argument

# CRS inferred from input and changed to 5070
aoi_ext("Fort Collins", wh = c(20000, 10000)) %>% 
  discritize(out_crs = 5070)

# CRS inferred
t1 = aoi_ext("Fort Collins", wh = c(20000, 10000), crs = 5070, bbox = TRUE) %>% 
  discritize(out_crs = 5070)
t1$extent

# Bad, assumed EPSG
t2 = c(-781506.2, 1976590.7, -740107.0, 2000454.0) %>% 
  discritize(out_crs = 5070)
t2$extent

# Explicit CRS named
t3 = c(-781506.2, 1976590.7, -740107.0, 2000454.0)  %>% 
  discritize(in_crs = 5070, out_crs = 5070)
t3$extent

SpatRaster

If you want a gridded object for future processing, you can set spatrast = TRUE.

aoi_ext("Fort Collins", wh = c(20000, 10000)) %>% 
  discritize(spatrast = TRUE)

Passing Other objects.

Defining and extent for a POINT and wh argument is only one option that can be discritized. Others include all AOI outputs, sf objects and both SpatVect and SpatRast objects.

AOI BBOX returns

geocode(c("Fort Collins", "Boulder"), bbox = TRUE) %>% 
  discritize(spatrast = TRUE, dim = c(10,10))

AOI fiat boundaries

(really this applies to any sf object!)

aoi_get(state = "OR") %>% 
  discritize(spatrast = TRUE, dim = c(5,5))

Other Spat* object

(r <- terra::rast(system.file("ex/elev.tif", package="terra")) )

# SpatRast
discritize(r, spatrast = TRUE)

# SpatVect
terra::vect(system.file("ex/lux.shp", package="terra")) %>% 
  discritize(spatrast = TRUE)

mikejohnson51/AOI documentation built on Nov. 22, 2023, 6:05 p.m.