fast_extract: Fast extract
In prioritizr/prioritizrutils: Utilities for prioritizr

Description Usage Arguments Details Value See Also Examples

Extract data from a Raster-class object from a Spatial-class object using performance enhancing tricks.

## S4 method for signature 'Raster,SpatialPolygons'
fast_extract(x, y, fun = mean,
  velox = requireNamespace("velox", quietly = TRUE), ...)

## S4 method for signature 'Raster,SpatialLines'
fast_extract(x, y, fun = mean, ...)

## S4 method for signature 'Raster,SpatialPoints'
fast_extract(x, y, fun = mean, ...)

`x`	`Raster-class` object.
`y`	`Spatial-class` object.
`fun`	`function` used to summarise values. Defaults to `sum`. Note that this only used when `x` is a `SpatialPolygons-class` or a `SpatialLines-class` object. This function must have an "na.rm" argument.
`velox`	`logical` should the `velox` be used for geoprocessing? Defaults to `TRUE` if the package is installed. Note that this only used when `x` is a `SpatialPolygons-class` object.
`...`	additional arguments passed to `extract`.

Spatial analyses will be conducted using the velox package if it is installed. Additionally, multiple threads can be used to speed up computation using the set_number_of_threads function.

data.frame, matrix, or list object depending on the arguments.

extract, VeloxRaster_extract.

# load data
data(sim_pu_polygons, sim_features)

# we will investigate severals ways for extracting values from a raster
# using polygons. Specifically, for each band in the raster,
# for each polygon in the vector layer, calculate the average
# of the cells that are inside the polygon.

# perform the extraction using the standard raster::extract function
system.time({result <- fast_extract(sim_features, sim_pu_polygons)})

# perform extract using the fast_extract function augmented using the
# "velox" package
system.time({result <- fast_extract(sim_features, sim_pu_polygons,
                                    velox = TRUE)})

# perform extract using the fast_extract function with "velox" package
# and using two threads for processing. Note that this might be slower
# due to overheads but should yield faster processing times on larger
# spatial data sets
set_number_of_threads(2)
system.time({result <- fast_extract(sim_features, sim_pu_polygons,
                                    velox = TRUE)})
set_number_of_threads(1)