Gepafer

Does the same thing as gepaf (which stands for Google Encoded Polyline Algorithm Format) but in a more efficient way, using Rust.

The name is still subject to discussion but "gepafer" can be understood as "gepaf, faster".

Motivation

Some packages, such as osrm, need to encode / decode coordinates in order to communicate with the OSRM API.

There are already some R packages that provide this functionality:

• The gepaf package is a great package (disclaimer: I'm its co-author with @rcarto), but it is written in pure R and it is relatively slow.
• The googlePolylines package is fast (it uses C++ under the hood) but, on 10/10/2024, the CRAN announced that the package will be archived on 24/10/2024 if nothing is done by its maintainer.

This package aims to provide the same functionality as gepaf while being as fast as googlePolylines, using Rust.

Moreover, it's a great opportunity to learn how to write R packages with Rust. To do so, gepafer is using the extendr Rust crate and the rextendr R package (but other alternatives such as savvy exist and could be considered in the future).

Installation

You can install the development version of gepafer from GitHub with:

# install.packages("remotes")
remotes::install_github("mthh/gepafer")

You will need the Rust toolchain to compile the Rust code. The Minimum Supported Rust Version (MSRV) is 1.61.0.

The package is not yet on CRAN.

Usage

This packages aims to be a drop-in replacement for the gepaf package. It provides two functions, encode_polyline and decode_polyline, that do the same thing as the functions with the same name in the gepaf package.

Encoding

The encode_polyline function takes a data frame with two columns, lat and lon, and returns an encoded polyline.

library(gepafer)

coords <- data.frame(lat = c(38.5, 40.7, 43.252), lon = c(-120.2, -120.95, -126.453))
encpoly <- encode_polyline(coords, factor=5)
encpoly
#> [1] "_p~iF~ps|U_ulLnnqC_mqNvxq`@"

Decoding

The decode_polyline function takes an encoded polyline and returns a data frame with two columns, lat and lon.

library(gepafer)

coords <- decode_polyline(enc_polyline = "_p~iF~ps|U_ulLnnqC_mqNvxq`@", factor=5)
coords
#>      lat      lon
#> 1 38.500 -120.200
#> 2 40.700 -120.950
#> 3 43.252 -126.453

License

GPL-3 (as the gepaf package)