View Package Documentation

The vectra package runs dplyr-style queries on datasets larger than RAM. It is backed by a pull-based columnar execution engine written in C11, so a query streams through the data one batch at a time and peak memory stays bounded regardless of how large the file on disk is.

The familiar verbs are all there: filter(), select(), mutate(), group_by(), and summarise(), along with the left_join(), inner_join(), and fuzzy_join() family, arrange(), the slice_*() helpers, and window functions such as row_number(), rank(), lag(), cumsum(), and ntile(). Execution is lazy: verbs build a plan that explain() can print, and collect() or collect_chunked() pulls the result when you ask for it.

Data lives in a custom on-disk columnar format (.vtr). vectra also reads and writes GeoTIFF, including tiled and BigTIFF layouts, and a tiled raster format (.vec) with overview pyramids and time cubes for larger-than-RAM raster work. Vector operations including spatial transforms, point-in-polygon and nearest-feature joins, select-by-location, clip, erase, dissolve, rasterization, polygonization, and contouring stream through sf. Raster operations including zonal statistics, focal windows, terrain derivatives, warp resampling and reprojection, polygon masking, map algebra, and mosaicking run in native C over the tiled format, one tile at a time.

Available on CRAN under MIT license.

install.packages("vectra")

Citation

Gilles Colling

PhD student at University of Vienna. Physicist turned ecologist. R packages, spatial statistics, and computational ecology.