*********************************
*** BMS TOOLBOX FOR MATLAB ******
***   version 0.2 (beta) ********
***   build 2011-01-09   ********
*********************************


__ ABOUT ________________________________________________________________

BMS toolbox for Matlab allows to perform Bayesian Model Averaging (BMA) 
with the R-library 'BMS' out of Matlab. 

The toolbox provides the core functionality to call an R session in the 
background that actually runs the BMS package

For more information, refer to http://bms.zeugner.eu/matlab/



__ INSTALLATION  & REQUIREMENTS _________________________________________

BMS toolbox for Matlab will work only on Windows (XP, NT, Vista, 7)
platforms with Matlab versions up from version 6.5

Note: The Operating system to have R installed as well as the the 
R (D)COM Server for Windows (http://rcom.univie.ac.at)
However that can be complicated.
It is far easier to install RAndFriends that contains 
both, even if that involves a quite large download and isntallation

1) Download RAndFriends setup from http://rcom.univie.ac.at/ 
    => 'Downloads' (this may take a while)
2) Install RAndFriends by double-clicking the downloaded file 
   (may take some minutes)
3) Save this toolbox in a convenient location on your hard drive



__ USING BMS ______________________________________________________

In order to do Bayesian Model Averaging with BMS toolbox for Matlab,
you have to load the library:
1) open Matlab
2) navigate your current directory to the toolbox location 'BMS_matlab'
3) execute the following command
    libraryBMS;
4) Now (maybe after some final installation) the BMS pacckage is up 
    and running
    
For a start try the following two possibilities:
* look at demo.m and its commands
* execute the following commands:+
_________________________
load datafls;
bmsGUI('datafls');
__________________________

Note that this Matlab toolbox provides just the core functionality 
of what is possible within R
It is thereforoe advised to perform any serious BMA work within R

__ AUTHOR & LICENCE _______________________________________________

This toolbox was written by Stefan Zeugner 2010 (http://www.zeugner.eu)
and is licenced under Creative Commons Attribution 3.0
( http://creativecommons.org/licenses/by/3.0/ )
Basically, anybody is free to use the included Matlab code for any purpose, 
and create derivative works etc.However, it is required that published computer code 
(in its comments section) or any scientific text give due reference to the website 
http://bms.zeugner.eu 
Please keep in mind that BMS toolbox for Matlab is still beta. 
Under no circumstances, the author will assume any liabilities for damages caused by using BMS.



__ TROUBLESHOOTING ________________________________________________

Q: In Matlab, the following appears repeatedly: 
  'BMS library seems to be not installed. Trying to install...':

A: for some reason Matlab does not succeed in installing the BMS package within R
  Open the R GUI (by Start->Run..., type RGui and press Enter) 
  in the menu choose 'Packages' -> 'Install Package(s)...',
  choose you mirror, scroll down to BMS and install

If you cannot find BMS then you probably have an older R version.
Try to install with the instructions from http://bms.zeugner.eu/getBMS
 __________________________________________________________________
Q: On executing libraryBMS in Matlab, an error message appears complaining 
  that R or Matlab has no write access to something
  
A: you have the wrong R version in your Windows system Path 
  check http://www.computerhope.com/issues/ch000549.htm
  adjust the Path to the newest R Version and restart Matlab
 __________________________________________________________________

Note: If you try to use the toolbox without RAndFriends, yoo might be 
experiencing problems with the interplay of the D-COM and Matlab