Road Network Projection R package roads version 1.2.1 Iterative least cost path and minimum spanning tree methods for projecting forest road networks. The methods connect a set of target points to an existing road network using 'igraph' <https://igraph.org> to identify least cost routes. The cost of constructing a road segment between adjacent pixels is determined by a user supplied weight raster and a weight function; options include the average of adjacent weight raster values, and a function of the elevation differences between adjacent cells that penalizes steep grades. These road network projection methods are intended for integration into R workflows and modelling frameworks used for forecasting forest change, and can be applied over multiple time-steps without rebuilding a graph at each time-step. Software Depends: R (>= 2.10) Imports: dplyr, igraph (>= 2.0.3), data.table, sf, units, rlang, methods, tidyselect, terra Suggests: testthat (>= 2.1.0), knitr, rmarkdown, viridis, bench, gdistance Josie Hughes <josie.hughes@ec.gc.ca> Comprehensive R Archive Network (CRAN) Sarah Endicott [aut] (ORCID: <https://orcid.org/0000-0001-9644-5343>), Kyle Lochhead [aut], Josie Hughes [aut, cre], Patrick Kirby [aut], Her Majesty the Queen in Right of Canada as represented by the Minister of the Environment [cph] (Copyright holder for included functions buildSimList, getLandingsFromTarget, pathsToLines, plotRoads, projectRoads, rasterizeLine, rasterToLineSegments), Province of British Columbia [cph] (Copyright holder for included functions getGraph, lcpList, mstList, shortestPaths, getClosestRoad, buildSnapRoads) Apache License (>= 2) 2026-01-16 application/tgz https://CRAN.R-project.org/package=roads doi:10.32614/CRAN.package.roads