Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/EhrmannS/geometr

Generate and modify interoperable geometric shapes
https://github.com/EhrmannS/geometr

geometric-shapes getters interoperability setters spatial-data

Last synced: 3 months ago
JSON representation

Generate and modify interoperable geometric shapes

Awesome Lists containing this project

README 

        
---

output: github_document

---



```{r, include = FALSE}

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)

```

# geometr 



[![R-CMD-check](https://github.com/EhrmannS/geometr/workflows/R-CMD-check/badge.svg)](https://github.com/EhrmannS/geometr/actions)

[![Coverage

Status](https://codecov.io/gh/EhrmannS/geometr/branch/master/graph/badge.svg)](https://codecov.io/github/EhrmannS/geometr?branch=master)

[![CRAN\_Status\_Badge](http://www.r-pkg.org/badges/version/geometr)](https://cran.r-project.org/package=geometr)

[![](http://cranlogs.r-pkg.org/badges/grand-total/geometr)](https://cran.r-project.org/package=geometr)



## Overview



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



  - produce an identical output for the same metadata of different classes (via so-called getters) and

  - use an identical input to write to various classes that originally require different input (via so-called setters).



## Installation



1) Install the official version from CRAN:



```{r, eval=FALSE}

install.packages("geometr")

```



or the latest development version from github:



```{r, eval=FALSE}

devtools::install_github("EhrmannS/geometr")

```



2) The [vignette](https://ehrmanns.github.io/geometr/articles/geometr.html) 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](https://en.wikipedia.org/wiki/List_of_geometers)\!



## Examples



Create a `geom`



```{r}

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).



```{r}

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.



```{r}

# 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.



```{r}

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.



```{r, fig.height=3}

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`.



```{r, fig.height=3}

myAttr <- "NWBIR74"

geo_vis(!!paste0("NC - ", myAttr) := nc_geom, fillcol = !!myAttr)

```



The `@window` slot of a modified `geom` is by default automatically updated.



```{r, fig.height=3.5}

geo_vis(`Currituck` = currituck, linecol = "fid")

```



Finally, cast a `geom` to another type simply by providing it in `anchor` of the respective type



```{r, fig.height=3.5}

boundPoints <- geo_point(crds = geomio::getPoints(currituck))

geo_vis(`Currituck - boundary vertices`= boundPoints, linecol = "fid")

```