https://github.com/hypertidy/fasterize
High performance raster conversion for modern spatial data 🚀🌏▦
https://github.com/hypertidy/fasterize
r raster rcpp rcpparmadillo rstats sf spatial
Last synced: 6 months ago
JSON representation
High performance raster conversion for modern spatial data 🚀🌏▦
- Host: GitHub
- URL: https://github.com/hypertidy/fasterize
- Owner: hypertidy
- License: other
- Created: 2017-03-09T22:50:54.000Z (over 8 years ago)
- Default Branch: main
- Last Pushed: 2025-03-08T21:14:18.000Z (7 months ago)
- Last Synced: 2025-03-31T16:13:53.545Z (6 months ago)
- Topics: r, raster, rcpp, rcpparmadillo, rstats, sf, spatial
- Language: C++
- Homepage: https://hypertidy.github.io/fasterize/
- Size: 1.08 MB
- Stars: 181
- Watchers: 9
- Forks: 15
- Open Issues: 28
-
Metadata Files:
- Readme: README.Rmd
- Changelog: NEWS.md
- Contributing: CONTRIBUTING.md
- License: LICENSE
- Code of conduct: CODE_OF_CONDUCT.md
- Codemeta: codemeta.json
Awesome Lists containing this project
README
---
output:
github_document:
html_preview: FALSE
---
```{r, setup, echo = FALSE, message = FALSE}
knitr::opts_chunk$set(
comment = "#>",
tidy = FALSE,
error = FALSE,
fig.width = 7,
fig.height = 4.5,
fig.path = 'vignettes/readme-',
cache=FALSE)
```
# fasterize
Fast polygon-to-raster conversion, burn polygon shapes and/or values into pixels.
[](https://github.com/hypertidy/fasterize/actions/workflows/R-CMD-check.yaml)
[](http://www.repostatus.org/#active)
[](https://badges.mit-license.org/)
[](https://CRAN.R-project.org/package=fasterize)
[](https://www.r-pkg.org/pkg/fasterize)
[](https://app.codecov.io/gh/hypertidy/fasterize)
**fasterize** is a high-performance replacement for the `rasterize()` function
in the [**raster**](https://cran.r-project.org/package=raster) package.
Functionality is currently limited to rasterizing polygons in [**sf**](https://cran.r-project.org/package=sf)-type
data frames.
## Installation
Install the current version of **fasterize** from CRAN:
```{r eval = FALSE}
install.packages('fasterize')
```
Install the development version of **fasterize** with [**devtools**](https://cran.r-project.org/package=devtools):
```{r eval = FALSE}
devtools::install_github("hypertidy/fasterize")
```
**fasterize** uses [**Rcpp**](https://cran.r-project.org/package=Rcpp) and thus requires a compile toolchain to install from source.
Testing (and for normal use of sf objects) requires [**sf**](https://cran.r-project.org/package=sf), which requires GDAL, GEOS, and PROJ to be installed.
## Usage
The main function, `fasterize()`, takes the same inputs as `raster::rasterize()` but currently has fewer options and is
is limited to rasterizing polygons.
A `raster()` and `plot()` methods for rasters are re-exported from the [raster package](https://cran.r-project.org/package=raster).
```{r example-1, message=FALSE}
library(raster)
library(fasterize)
library(wk)
library(fasterize)
p123 <- c(paste0("POLYGON ((-180 -20, -140 55, 10 0, -140 -60, -180 -20),",
"(-150 -20, -100 -10, -110 20, -150 -20))"),
"POLYGON ((-10 0, 140 60, 160 0, 140 -55, -10 0))",
"POLYGON ((-125 0, 0 60, 40 5, 15 -45, -125 0))")
pols <- data.frame(value = seq_along(p123), geometry = wk::as_wkt(p123))
ex <- as.numeric(wk_bbox(pols))[c(1, 3, 2, 4)]
r <- raster::raster(raster::extent(ex), res = 1)
r <- fasterize(pols, r, field = "value", fun="sum")
plot(r)
```
## Performance
Let's compare `fasterize()` to `terra::rasterize()`:
```{r benchmark, cache=TRUE}
pols_t <- terra::vect(p123)
pols_t$value <- 1:3
#pols_r <- as(pols_t, "Spatial")
tr <- terra::rast(r)
bench <- microbenchmark::microbenchmark(
# rasterize = r <- raster::rasterize(pols_r, r, field = "value", fun="sum"),
terrarize = tr <- terra::rasterize(pols_t, tr, field = "value", fun = "sum"),
fasterize = f <- fasterize(pols, r, field = "value", fun="sum"),
unit = "ms"
)
print(bench, digits = 3)
```
How does `fasterize()` do on a large set of polygons? Here I download the IUCN shapefile for the ranges of all terrestrial mammals and generate
a 1/6 degree world map of mammalian biodiversity by rasterizing all the layers.
(this doesn't work anymore because the source data is gone, left as a record 2024-09-25).
```{r download, eval=FALSE, cache=TRUE}
if(!dir.exists("Mammals_Terrestrial")) {
download.file(
"https://s3.amazonaws.com/hp3-shapefiles/Mammals_Terrestrial.zip",
destfile = "Mammals_Terrestrial.zip") # <-- 383 MB
unzip("Mammals_Terrestrial.zip", exdir = ".")
unlink("Mammals_Terrestrial.zip")
}
```
```{r so-damn-fast, cache=FALSE, eval=FALSE}
mammal_shapes <- st_read("Mammals_Terrestrial")
mammal_raster <- raster(mammal_shapes, res = 1/6)
bench2 <- microbenchmark::microbenchmark(
mammals = mammal_raster <- fasterize(mammal_shapes, mammal_raster, fun="sum"),
times=20, unit = "s")
print(bench2, digits=3)
par(mar=c(0,0.5,0,0.5))
plot(mammal_raster, axes=FALSE, box=FALSE)
```
#> Unit: seconds
#> expr min lq mean median uq max neval
#> mammals 0.847 0.857 0.883 0.886 0.894 0.963 20
## About
**fasterize** was developed openly at [EcoHealth Alliance](https://www.ecohealthalliance.org/) under the USAID PREDICT project by Noam Ross. The repository for hosting fasterize was taken over by Michael Sumner in December 2022, and was later migrated from Github 'ecohealthalliance/fasterize' to https://github.com/hypertidy/fasterize in March 2025.
Please note that this project is released with a [Contributor Code of Conduct](CODE_OF_CONDUCT.md). By participating in this project you agree to abide by its terms.
[](https://www.ecohealthalliance.org/)
[](https://ohi.vetmed.ucdavis.edu/programs-projects/predict-project)