File	Date	Author	Commit
R	2013-06-26	Dominik Wabersich	[b61634] removed unnecessary lambda restriction (<1)
data	2013-03-05	Dominik Wabersich	[4f892b] Added likelihood functions for optim use
man	2013-05-22	Dominik Wabersich	[4f4523] minor doc fix
src	2013-03-04	Dominik Wabersich	[682ff3] fixed indexing bug, which caused random behavior
DESCRIPTION	2013-02-16	Dominik Wabersich	[5c4e3a] Initialized repo
NAMESPACE	2013-03-05	Dominik Wabersich	[4f892b] Added likelihood functions for optim use
README	2013-06-14	Dominik Wabersich	[6af21c] updated README
cleanup	2013-02-16	Dominik Wabersich	[5c4e3a] Initialized repo
configure.ac	2013-02-16	Dominik Wabersich	[5c4e3a] Initialized repo
configure.win	2013-02-16	Dominik Wabersich	[5c4e3a] Initialized repo

Read Me

RAlcove R package
=================
The RAlcove package is an extension for R, which provides R with functions
for the ALCOVE (Attention Learning Covering Map) model. The ALCOVE model is
a cognitive model that models categorization in humans by combining an
exemplar-based approach with error-driven learning.

Besides the main alcove() function that computes the model predictions for
given data, the package provides a likelihood function, which can be used
in combination with R's optim() function, to get estimates for the model
parameters.

Using the package
-----------------
::

  alcove(stim, truecat, learn, 
         alpha_init, omega_init, h, 
         lam_o, lam_a, c, phi, 
         q, r)

*the data*:

- stim: a vector containing the stimulus numbers.
- truecat: a vector containing the true categorization.
- learn: a vector containing 1 if feedback is given, 0 otherwise.

*necessary model variables*:

- alpha_init: a vector with the initial alpha values.
- omega_init: a matrix with the initial omega values.
- h: a matrix with the psychological stimulus dimension for every stimulus.

*model parameters*:

- lam_o: omega lambda learning parameter.
- lam_a: alpha lambda learning parameter.
- c: specifity parameter.
- phi: probability mapping constant.

- q: further parameter of the model, usually fixed at 1.
- r: further parameter of the model, usually fixed at 1.

See the R manual pages for more details, examples and usage of the other functions.

Please note
-----------
Copyright (C) 2013 Dominik Wabersich <dominik.wabersich@gmail.com>,
Michael Lee <mdlee@uci.edu> and Joachim Vandekerckhove <joachim@uci.edu>

License
-------
http://www.r-project.org/Licenses/GPL-2