spatstat is an R package for analysing
spatial point pattern data (and other kinds of spatial data).
See the website www.spatstat.org
or read the book.
This repository holds a copy of the current development version
of spatstat.
This development version is more recent than the official release
of spatstat on CRAN. Each official release of spatstat has a
version number like 1.2-3 while the development version has a
version number like 1.2-3.004 (which R recognises as a later version).
Official releases occur every 8 weeks
(the minimum time permitted by CRAN policies) while the development code
is updated almost every day.
For the most recent official release of spatstat,
see the CRAN page.
Recently we have started the process of splitting spatstat into several
packages (as mandated by CRAN, because spatstat is very large).
Your existing code will still work:
typing library(spatstat) will still give you
access to all the functions in spatstat that you know from previous versions.
However, messages from R about the installation and loading of the package
will now show that spatstat consists of several pieces.
Currently there are four pieces:
-
spatstat: contains the main functionality of thespatstatfamily. -
spatstat.data: contains the datasets for thespatstatfamily. The current development version ofspatstat.datais here. -
spatstat.utils: utility functions originally included inspatstatwhich are now accessible as a separate package. The current development version ofspatstat.utilsis here. -
spatstat.sparse: functions for linear algebra and sparse arrays, originally included inspatstatwhich are now accessible as a separate package. The current development version ofspatstat.sparseis here.
When you type library(spatstat) this will load
the main spatstat library
and the spatstat.data library,
and will also import the spatstat.utils and spatstat.sparse libraries.
Importing means that the spatstat.utils and spatstat.sparse functions
can be used by spatstat but cannot be accessed by
the user. To access these utility functions directly, you need to type
library(spatstat.utils) and library(spatstat.sparse) respectively.
There are also extension packages which provide additional capabilities and must be loaded explicitly when you need them. Currently there are four extension packages:
. spatstat.local for local model-fitting,
. spatstat.Knet for analysing point patterns on a network of lines,
. spatstat.gui
for interactive graphics (containing the functions iplot and istat that
were previously part of spatstat),
. spatstat.sphere for analysing point patterns on a sphere (package is still in preparation).
To install the official release of spatstat from CRAN, start R and type
install.packages('spatstat')The easiest way to install the development version of spatstat
from github is through the remotes package. Start R and type
require(remotes)
install_github('spatstat/spatstat.utils')
install_github('spatstat/spatstat.data')
install_github('spatstat/spatstat')If you don't have remotes installed you should first run
install.packages('remotes')For Windows users: installing via remotes requires that RTools is installed
and working. If this is not an option for you, you can also install the
development version by downloading the zip file from
appveyor
(bottom of the page) and installing the package from the downloaded file (which
will be much, much faster than the remotes method).
Users are encouraged to report bugs here. Go to issues in the menu above, and press new issue to start a new bug report, documentation correction or feature request.
Please do not post questions on the Issues page, because it's too clunky for correspondence.
For questions about spatstat, first check
the question-and-answer website stackoverflow.
If your question is not listed,
you can either post your question at stackoverflow, or
email the authors.
Feel free to fork spatstat, make changes to the code,
and ask us to include them in the package by making a github pull request.
However, this repository is only a copy of the development code, so your pull request may not be implemented until the next official release.

