R-beta: R-0.61.3 for Win95/NT available

[guido@sirio.stat.unipd.it]

I have just put on ftp://sirio.stat.unipd.it/pub/R a new version 
of R for Win95/NT.
The main difference concerns dynamic linking (see below).
In addition, the binary distribution contains many of the 
available packages (all the ones that I have installed on
my Unix machine).
In the afternoon, I will upload the files on CRAN.
g.m.

This is the README file:
rw0613b.zip is  a new binary distribution of R (based on the last "stable"
version, 0.61.3) for Win95 and WinNT. It is compiled using  egcs-minw32.
rw0613s.zip contains the MsWindows sources.

News
(a) Now, a dll can call a routine in the main executable. This
    means that call_S, R_alloc,... work. Try, for example,
    demo(dyn.load).
    All R.exe (non static) functions and all global variables
    are exported.
    In this way, more packages can be ported and furthermore
    the size of the dll cuts down.
(b) I fixed a nasty bug which made impossible to set the
    memory size. Now, R -v VVVV -n NNNN works.
(c) There is a new command line switch -home aDir which
    sets for the current run the HOME directory.
    Workspace and user RProfile file are  searched in HOME.
    In addition, now tempfile() try to use the environmental variables 
    TMP, TEMP and HOME (in this order).
(d) The following packages are included in the binary distribution:
      acepack    ace() and avas() for selecting regression transformations 
      base       The R base package 
      bootstrap  Functions for the Book "An Introduction to the Bootstrap" 
      class      Functions for classification 
      ctest      Classical Tests 
      eda        Exploratory Data Analysis 
      gee        Generalized Estimating Equation models  
      integrate  numerical integration 
      leaps      Regression subset selection 
      locfit     Local Regression, Likelihood and Density Estimtion. 
      mva        Classical Multivariate Analysis 
      polymars   (polychotomous) regression based on  
                 Multivariate Adaptive Regression Splines 
      polynom    A collection of functions to implement a class for  
		 univariate polynomial manipulations 
      spatial    functions for kriging and point pattern analysis 
      splines    Regression Spline Functions and Classes 
      stepfun    Step Functions, including Empirical Distributions 
      survival4  Survival analysis [needs library(splines)] 

    Of course, I cannot test all the functions. So, don't flame me
    if something doesn't work.
    

Installation (binary distribution)
Just unzip the archive. You will get a new directory (rw0613).
The executable is bin/R.exe.

Installation (sources)
First of all, you have to download and untar the complete sources.
Then, in the directory R-0.61.3, unzip rw0613s.zip.
To rebuild the system using egcs-mingw32 go to the src/gnuwin32
directory and type 
> make
This will build the main executable (in R-0.61.3/bin) and
the standard packages: base, eda, mva, stepfun.

Installing other packages
Uncompress the package you want in R-0.61.3/src/library.
Then, from the gnuwin32 directory, 
> make -f MakePkg -C ../library/xxx PKG=xxx [HAVEG77=NO] [DLLNAME=yyy]
will try to build the package xxx.
The HAVEG77=NO options must be used if:
(a) you don't have or don't want to use g77;
or
(b) the package has a src-c directory but not a src one.
The option DLLNAME=yyy must be used if the package requires a
dll with a name different from xxx.dll. For example, the dll
of integrate is adapt.dll.
If you don't need the two flags, you can use the short command
> make pkg-xxx
Compiler flags must be setted in MakeDll. Look to this file
for some example.


Help files
There are no support for generating the help files.
You have to download the pre-formatted one by any CRAN site or
to build them yourself on a Unix machine.

Tools
To rebuild the system and to install other packages, you need
not only a compiler but also some Gnu tools. The ones that I used,
coming from DJGPP, are  in djtools.zip.
You must uncompress this file from C:. It will expand
under usr/bin and usr/home.
To use the tools usr/bin must be in the PATH.
In djtools.zip, a version of bash is included.
I am not sure that you need it but I did all my work 
using it. To use bash, the environmental variable HOME
should point to /usr/home. You have to start bash, in a console
or a MsDos windows, with
> bash -login
If you install djtools.zip, please use the make in usr/bin not
the one of your compiler.


guido masarotto (22/05/1998)
guido at sirio.stat.unipd.it
http://sirio.stat.unipd.it




-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._