Changes in version 2.1                         

    o   Corrected bug in gw in relation of optimization methods.

    o   residuals.gw S3 method includes now the possibility of parallel
	computing in order to construct the simulated envelope.

                  Changes in version 2.0 (2015-04-21)                   

    o   GWRM.fit is now gw. The default algorithm to find ML-estimates
	is L-BFGS-B after nlm and Nelder-Mead previous optimizations;
	it provides SE estimates of all the parameters of the model.

    o   gw with k=1 argument corresponds to the classical Waring
	regression model.

    o   GWRM.display is now the S3 method summary.gw.

    o   GWRM.add is now included in the new S3 method add1.gw and may
	be used through step.

    o   GWRM.stats is now the S3 method partvar.gw.

    o   A residuals.gw S3 method has been included to validate the fits
	with the analysis of residuals.

    o   NEWS and CITATION files included.