Using R in QGis processing

Although R is essentially a set
of statisticals, its use has been extended to a number of
areas, such as natural language processing. This
library
possesses
also
solic tools of spatial analysis,
including analysis of the distributions of points and the Bayesian
geostatical modelling. It is possible to read and manage a
variety
of vector and raster data, including shapefiles, NetCDF and GDAL.

Traditional GIS programs often
have
been limited by   their spatial statistics and their analytical capacity were relatively
minor
, including a small range of integrated
functions, forcing users to use other platforms for advanced analysis,
modelling and simulation. ESRI made the
acquisition of the MapStudio program from the University of Minsk, which from  2001 has become the Geostatistical Analyst
extension of ArcGis .

With R, many popular statistical proceedings
and
more advanced analyses, including a variety
of simulation applications , can be
applied
directly
in
tools such as QGIS.

How to install R for QGis.

No need to rush to the QGis processing
panel to find R. It is not part of the processing providers included
with the QGis installation.

You have to download
and
install R on your computer before configuring QGis.

Here is the link to download R: https://cran.r-project.org/bin/windows/base/

Once downloaded , double click on
the file R-3.3.2-win.exe to
install R on your computer .

The program installs R in the directory C:
\ Program Files \ R \ R-3.3.2

How to configure the
QGis
Processing
Options

To configure R , open the window
Options
treatments (treatments-> Options …)

Open Treatmentproviders-> R scripts

Check the Activate box to
activate the option R and for it to appear in
the panel Processing .

If you have a
64bits Windows version, you can check
the box Use the 64-bit version . If you
have
a
32-bit post, only the 32-bit version of R is installed
. If you have a
64-bit computer, both versions (32 and 64 bits) are installed
. If you do not check the box you will
use the same 32-bit version if your
PC is a 64 bit.

You have to configure the
directory with the library R for QGis. This
directory , rlibs lies in
C: / User / your username /.qgis2/processing

if it does not exist , create
a
repertoire
rscripts beside rlibs
and
indicate this repertoire in
As
Scripts Directory R.

Point the R directory to the directory installation
, default C: \ Program Files \ R \
R-3.3.2

Click OK to close the window Treatments
Options.

You have now
the provider R that appears in the processing panel, with
two sub-items:

The first Create new R script
opens
an
editing window or you
can
type a Python script. If you have inputs
to set for execution , this is the moment to do it.

The second , Get R scripts from on-line scripts
collection
, allows you to use Python scripts ready
to use with the adequate processing windows already
configured .

Let’s see an example
. You want to carry
out an histogram R for a field of a
 vector layer .

Double- click on Get R
scripts from on-line scripts collection
to open the window Getscriptsandmodels:

Open item Notinstalled :   

Search the tool Histogram
, check the corresponding box and click
OK.

With version 2.18.2 which I
use, the processing window will not update immediately. If you open
the window again Getscriptsandtemplates you
will
see a new item in Updateavailable (in the case of
the histogram) or item Installed .
 

When you close
this
window
, the R processing window in the panel is updated,
and your new tool is
now
available
under UserR script-> Histogram

If
you double click on the tool , you will get the settings window
that opens:  

When you click
on Run , the processing is performed :

If you are not familiar with R, this
is the easiest way to start using the tools that R puts at your disposal.

If you make a
right
click on the tool in the processing panel, you
can
select
the
option Editthe script
to see the construction of the Python script:

Here it is!. We have
half-opened a door for you to discover
the possibilities that R can bring you.

Leave a Reply

Your email address will not be published.