Version 1.1-1
* Set default alpha for rlga to 0.9 (was 0.5)
* Updated reference for rlga
* Added a new data set (corridorWalls), added a man file, and included an 
example for rlga using this data
* Change now perform orthogonal regression (both non-robust and robust) by
setting k = 1.
* Some behind-the-scenes changes to the lga object model (biter, niter, k are
now attributes)

Version 1.1-0
* Added robust version of LGA (called rlga)
* Fixed and removed some functionality for pairs plots (no longer draws lines
when d > 2, which were mostly counter-intuitive)
* gap, lga and rlga are now generic methods (with defaults provided)
* Fixed a bug with scale() in lga etc.

Verion 1.0
* First release to CRAN
