R-gui is a development software project promoted by University of Trento,in particular by the informatic garrison of Trento and Rovereto.
The goal of the plan is the creation of a series of small programs that offer one graphical interface towards specific functionalities of R (http://www.r-project.org); not therefore one only generic interface to all R but many specific interfaces for “specific plans”.
“The increased” aim of the plan is the creation of with of instruments, documentation, examples etc. that facilitate the creation of these programs and the reuse of the job already made.
On this site you can find of documents that they illustrate the Motivations of this plan, the chosen Base software for development and the deploy, the Instruments Software of support to the plan.
The base motivations of this plan are substantially two:
This plan aimed therefore to indentificare of the specific customers, or the specific necessities, and developing of the “solutions aimed” of use much simple one and, not little important, a community of customers and persons of support, for which this situated one constitutes the point of encounter begins them.
So that the development of these solutions is not a sporadic event, the plan previews also the creation of presupposed (the software of development and support, documentation, examples, etc) that they render “easy” to continue and to expand the job made “in before struck”.
Beginning page---->
Inasmuch as they must be produced of the programs that are leaned to R, it is obvious that it serves a toolkit graphical, the bookcase R and something in order to connect them. As toolkit graphical it has been chosen GTK+ (http://www.gtk.org), like adhesive Python (http://www.python.org).
In order to connect Python and GTK+ the bookcase PyGTK (http://www.pygtk.org), is used in order to construct the interfaces uses Glade (http://glade.gnome.org) and in order to approach the definition of the interfaces from Python http://www.jamesh.id.au/software/libglade) is used libglade .
In order to connect Python to R, or better to the bookcase that supplies interpreter R, http://rpy.sourceforge.net) is used rpy .
Details
Prerequirement for the operation of the software vary, in the detail, to second of the platform (Linux vs. Windows).
Linux: For Linux, using the precompiled packages, are the usual differences due to the distributions. The distro of reference, because used from Luca Manini, it is Debian (http://www.debian.org).
Debian
The system of distribution of Debian is based on the usual rows .deb recovers to you via net (using programs like apt or synaptic), therefore enough to have the list of the deb. The more comfortable thing, in reality is knowledge which is the two or three package that they have, between their dependencies, all that that serves among other things (, in this way, would be banal to make a goal-package!).
There are some things important to remember:
in Debian many package are generic, in the sense that do not specify the version; package python ago the reference to the more recent version available. As an example, at the moment (june the 2006) version current of Python is the 2,4, but all the package that they depend on Python are not dawn to you, therefore on one “generic” machine even serves still also the 2.3.
The command apt-cache show <package> extension more “details” on the dependencies if the package it is not generic. As an example, for python-rpy he gives like dependencies single python (>= 2,3 << 2,4) and ptyhon2.3-rpy, but if specific ptyhon2.3-rpy, the true” packages are discovered “(es: python2.3-numeric, r-base-Core).
The command dpkg-depcheck <comando> executes <comando> under strace (discovering therefore to all the rows “loads to you”) and from this he deduces and extension the list of the necessary packages.
The command apt-cache depends <package-list> supplies the direct dependencies of one or more packages; as an example:
apt-cache depends python2.3-rpy python2.3-gtk2 libglade2-0
he gives a list that to me seems completes (for the deploy), therefore the installation of these three packages would have “to pull” all the rest.
However, on the whole the fundamental packages are:
python r-base-Core python-rpy libgtk2.0-0 python-gtk2 libglade2-0 python-glade2
Windows:
The procedure for the installation under Windows has been simplified thanks to the use of py2exe and InnoSetup
The first one allows “impacchettare” of the Python sources in a eseguibile Windows stand-halo that does not have need of an installation of Python in order to work but that it is leaned to a sure number of DLL and other rows that come but dstribuiti in one only directory; simplifying therefore a lot the installation. The second is a free software for the creation of programs of installation for Windows.
To this point the “packages” to install are reduced to three:
1) R: to be connected to http://www.r-project.org and to follow the link to the several one mirror of CRAN or to try a link direct like this
2)the bookcases of GTK+: to be connected to http://gladewin32.sourceforge.net/modules/news and to unload from using link “the Gtk+/Win32 Runtime Environment Installer 2.8.18-1 (.exe, 5.27M)”
E' not given to understand the logons between several the pieces of GTK well (they depend at least on who has pacchettizato what, where and when). One good source e' however:
http://www.pygtk.org/downloads.html
from which these can be unloaded two parts that, usual, lack and are instead indispensable:
http://ftp.gnome.org/pub/GNOME/binaries/win32/pygtk/2.10/pygtk-2.10.1-1.win32-py2.4.exe http://ftp.gnome.org/pub/GNOME/binaries/win32/pygobject/2.12/pygobject-2.12.1-1.win32-py2.4.exe
Beginning page---->