Package 'manby'

Title: Vector Tilesets Like 'tippecanoe', Several Times Faster
Description: Builds vector tilesets in 'PMTiles' and 'MBTiles' from sf data frames, 'DuckDB' tables, 'Arrow' tables and files, with the tiles that 'tippecanoe' makes from the same features. 'manby', a 'Rust' port of 'tippecanoe', is built into the package from CRAN and in 'webR', and elsewhere the package runs the 'manby' program, which it downloads from the package's release on first use.
Authors: Birk Skyum [aut, cre, cph], Mapbox Inc. [cph] (tippecanoe, which manby ports), Protomaps LLC [cph] (tippecanoe), The authors of the Rust crates that manby builds in [ctb, cph] (listed in inst/AUTHORS of the source package from CRAN)
Maintainer: Birk Skyum <[email protected]>
License: BSD_2_clause + file LICENSE
Version: 0.2.8
Built: 2026-11-14 14:11:06 UTC
Source: https://github.com/birkskyum/manby

Help Index


Runs manby

Description

Runs manby with arguments as on the command line, such as ⁠manby join⁠.

Usage

manby(...)

Arguments

...

The arguments.

Value

Nothing. If manby fails, an error of class manby_error, with its exit status in status: 100 when the tileset is complete only to a lower zoom, as for tippecanoe.

Examples

## Not run: 
manby("join", "-o", "map.pmtiles", "roads.pmtiles", "places.pmtiles")

## End(Not run)

The manby program

Description

manby_binary() gives the path of the manby program that the package runs: the one that the manby.binary option or the MANBY_BINARY environment variable names, or else the one that install_manby() downloaded. If there is neither, it calls install_manby().

Usage

manby_binary()

install_manby(version = own_version(), force = FALSE)

Arguments

version

The version of manby to download. The default is the package's own.

force

Download it again, even if it is in the cache.

Details

install_manby() downloads manby from the GitHub release of the package's version, checks it against the release's checksums, and keeps it in the package's cache, tools::R_user_dir(). manby runs on macOS, Linux and Windows.

Value

The path of the program.

Examples

## Not run: 
manby_binary()

## End(Not run)

Serves tilesets with a map of them

Description

Serves PMTiles archives, MBTiles databases or directories of tiles with ⁠manby serve⁠, until ⁠$stop()⁠ or the end of the R session. The map, at ⁠$url⁠, shows each tileset's layers on a basemap, with the attributes of the features clicked. A tileset made again while it is served shows when the map is reloaded.

Usage

serve(tilesets, port = 8080, host = "127.0.0.1", open = interactive())

Arguments

tilesets

The paths of the tilesets.

port

The port, or 0 for any free port.

host

The address to listen on.

open

Show the map, in the viewer pane or a browser. With manby built into the package, as from CRAN, it serves on threads of R's own process, until ⁠$stop()⁠ or the server is garbage collected.

Details

Each tileset's tiles and TileJSON can also go in a map of one's own, as can a PMTiles archive itself, at its file name after ⁠$url⁠.

Value

A manby_server, with url, the map's address, tilejson, the TileJSON address of each tileset by name, stop(), which ends it, and alive(), whether it serves.

Examples

## Not run: 
server <- serve("roads.pmtiles")
server$tilejson[["roads"]]
server$stop()

## End(Not run)

Makes a tileset

Description

Makes the tileset output from data, with the tiles that tippecanoe makes from the same features.

Usage

tile(data, output, ..., layer = NULL, force = TRUE, h3_sql = NULL)

Arguments

data

The features: a file name, a table, or a named list of layers.

output

The tileset to write, such as "roads.pmtiles".

...

tippecanoe's options. A named argument is a long option, so maximum_zoom = 14 is --maximum-zoom=14, drop_densest_as_needed = TRUE is --drop-densest-as-needed, and a vector repeats an option. Other arguments are passed as they are, such as "-zg".

layer

The name of the layer, for a single input. Without one, the layer is named after the file, or for a table after the tileset.

force

Overwrite an existing tileset.

h3_sql

Aggregates for the hexagons of --h3, as a named vector of DuckDB SQL, such as c(p90 = "quantile_cont(price, 0.9)"). DuckDB computes each over the points of each H3 cell, with its spatial and H3 extensions, and the hexagons get them as attributes of those names. The points of a Parquet file are taken to be in WGS 84.

Details

data is a file that manby reads, such as GeoJSON, FlatGeobuf, CSV, GeoParquet, GeoPackage or Shapefile, a file in another format that GDAL reads, or a table:

  • An sf data frame. One in a CRS other than WGS 84 or Web Mercator is reprojected to WGS 84 first.

  • A DuckDB table from dplyr, such as dplyr::tbl(con, "roads") or dplyr::tbl(con, dplyr::sql("SELECT ...")), which DuckDB writes to a temporary Parquet file, so it need not fit in memory.

  • An Arrow table or stream, or a data frame, with its geometry in a GeoArrow column or in a column named geometry in well-known binary.

A named list makes a layer of each of these. Other tables are written to a temporary Arrow file for manby to read.

A file in a format that GDAL reads and manby does not, such as KML, GPX, GML, OpenStreetMap (.osm.pbf), SQLite, a File Geodatabase (.gdb) or a spreadsheet, is read with sf, or with DuckDB's spatial extension if sf is not installed, and reprojected to WGS 84. The layer's name picks the source's layer of that name, if it has one, and otherwise the first.

Value

output, invisibly.

Examples

## Not run: 
nc <- sf::st_read(system.file("shape/nc.shp", package = "sf"))
tile(nc, "nc.pmtiles", layer = "counties", maximum_zoom = "g")

con <- DBI::dbConnect(duckdb::duckdb())
roads <- dplyr::tbl(con, dplyr::sql(
  "SELECT name, class, geometry FROM 'roads.parquet' WHERE class = 'motorway'"
))
tile(roads, "roads.pmtiles", maximum_zoom = 14, drop_densest_as_needed = TRUE)

tile(list(roads = roads, places = "places.parquet"), "map.pmtiles", "-zg")

tile(places, "places.pmtiles", maximum_zoom = 12, h3 = TRUE,
  h3_sql = c(p90 = "quantile_cont(price, 0.9)", kinds = "count(DISTINCT kind)"))

tile(list(points = "liechtenstein.osm.pbf", lines = "liechtenstein.osm.pbf"),
  "liechtenstein.pmtiles", maximum_zoom = 14, drop_densest_as_needed = TRUE)

## End(Not run)

The sizes of a tileset's tiles

Description

How big a PMTiles archive's, an MBTiles database's or a directory's tiles are at each zoom, as stored, and the largest tiles. With layers, also how many features of each layer the tiles at each zoom hold, which shows where dropping and coalescing took effect.

Usage

tile_stats(tileset, layers = FALSE, largest = 10)

Arguments

tileset

The path of the tileset.

layers

Count the features of each layer, from every tile.

largest

How many of the largest tiles to list.

Value

A list of data frames: zooms, with the tiles at each zoom and their total, mean, median, 99th percentile and largest size in bytes; largest, the largest tiles; and with layers, layers, the tiles and features of each layer at each zoom.

Examples

## Not run: 
tile_stats("roads.pmtiles", layers = TRUE)

## End(Not run)