knitr::opts_chunk$set( collapse = TRUE, comment = "#>", fig.path = "man/figures/README-", out.width = "100%" ) # calling any function from geometr for the first time creates an annoying # message about spatstat, which I can remove by calling it here without # including this chunk in the Rmd library(sf) library(geometr) nc_sf <- st_read(system.file("shape/nc.shp", package="sf"), quiet = TRUE) geomio::getFeatures(nc_sf)
The geometr
package provides tools that generate and process easily accessible and tidy geometric shapes (of class geom
). Moreover, it aims to improve interoperability of spatial and other geometric classes. Spatial
classes are typically a collection of geometric shapes (or their vertices) that are accompanied by various metadata (such as attributes and a coordinate reference system). Most spatial classes are thus conceptually quite similar, yet a common standard lacks for accessing features, vertices or the metadata. geometr
fills this gap by
providing tools that
1) Install the official version from CRAN:
install.packages("geometr")
or the latest development version from github:
devtools::install_github("EhrmannS/geometr")
2) The vignette gives an in depth introduction, explains the take on interoperability and discusses the spatial class geom
that comes with geometr
.
3) Have fun being a geometer!
Create a geom
library(geometr) # ... from other classes library(sf) nc_sf <- st_read(system.file("shape/nc.shp", package="sf"), quiet = TRUE) nc_geom <- geomio::as_geometr(x = nc_sf)
Metadata of different classes can be extracted in interoperable quality (i.e. the same metadata in different objects/classes have the same names and the same arrangement).
geomio::getExtent(x = nc_geom) geomio::getFeatures(x = nc_geom)
geometr
only knows the feature types point
, line
, polygon
and grid
(a systematic lattice of points).
In contrast to the simple features standard, there are no MULTI* features.
The way simple features have been implemented in R has the consequence that certain information can only be stored in a rather inefficient and cumbersome way.
Think of a set of polygons that has individual attributes at the POLYGON
level, but also attributes at the group (MULTIPOLYGON
) level, i.e., where all polygons of that group have the same value of the respective attributes.
It is impossible to store this information in a single sf
object, other than in a POLYGON
that contains the respective attributes as duplicates (which may be very inefficient in large objects).
The backbone of a geom
are three tables, one for points, one for features and one for groups of features, the latter two of which can be provided with ancillary information (attributes).
Each point is stored with a feature ID (fid
) that relates it to a feature (and its attributes) and each feature is stored with a group ID (gid
) that relates it to a group (and its attributes).
Eventually this results in a tidier data-structure with easier access than sp
or sf
objects and with higher versatility.
# when using the group = TRUE argument, the attributes of MULTI*-feature are # stored in the group attribute table of a geom nc_geom <- geomio::as_geometr(x = nc_sf, group = TRUE) currituck <- geo_filter(obj = nc_geom, gid == 4) geomio::getFeatures(x = currituck) geomio::getGroups(x = currituck) # and new attributes can be set easily, newTable <- data.frame(fid = c(1:108), attrib = rnorm(108)) (nc_geom <- geomio::setFeatures(x = nc_geom, table = newTable))
Moreover, a geom
also allows to store raster information in the same data format.
A raster is a systematic lattice of points/pixels with one or more layers, where each pixel exists in each layer, possibly with different values, i.e., from different attributes.
The raster attribute table (RAT) is basically already the groups attribute table of a geom
, where a group of pixels that have the same value are associated to attributes this class has in common.
In this case, however, the features (single pixels) don't have to be associated to a group ID, as their feature values are at the same time the group values.
str(geoms$polygon, max.level = 2) str(geoms$grid, max.level = 2)
geometr
comes with the visualise
function, which makes nice-looking diagnostic spatial plots, that show explicit values whenever possible.
For example, it does not create artificial (\"nice\") bins for the x and y axis or the values values shown in the plot, but provides the range of values explicitly.
Moreover, you can easily set plot titles without much effort.
geo_vis(`North Carolina` = nc_geom)
You can use quick options to scale plot parameters to the attributes of the geom
to modify which aspects of an object should be shown in the plot, for example when scaling fillcol
to NWBIR74
.
myAttr <- "NWBIR74" geo_vis(!!paste0("NC - ", myAttr) := nc_geom, fillcol = !!myAttr)
The @window
slot of a modified geom
is by default automatically updated.
geo_vis(`Currituck` = currituck, linecol = "fid")
Finally, cast a geom
to another type simply by providing it in anchor
of the respective type
boundPoints <- geo_point(crds = geomio::getPoints(currituck)) geo_vis(`Currituck - boundary vertices`= boundPoints, linecol = "fid")
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.