mgcv version 1.2 is on CRAN now. mgcv provides generalized additive models
and generalized additive mixed models with automatic estimation of the
smoothness of model components.
Changes in this version are:
* A new gam fitting method is implemented for the generalized case. It
provides more reliable convergence than the previous default, but can be
a little slower. See ?gam.method, ?gam.fit2 and ?gam.outer for details.
The old method is still available as an option.
* `gam' has acquired a new list argument `method' to cope with the
number
of fitting method options now available, and there has been some alteration
to the `control' argument.
* Any smoothers can now be used to construct nested models, including
tensor product smooths. See ?fixDependence and ?gam.side for details.
* By default all smooths are now parameterized to be centred, without
requiring additional constraints (this is automatic and applies also to
user defined smooths). The old behaviour is still available as an option.
See ?smoothCon for details. (This should be user transparent.)
* Smoothing parameter initialization has been modified for better
performance with tensor product smooths. See ?initial.sp.
* By default, tensor product smooths have been modified to use more
interpretable penalties. See ?te for details. This leaves smooths based
on the default basis unchanged, but improves the performance of smooths
based on other marginal bases.
* Examples of how to obtain variance estimates etc. efficiently, for any
quantities derived from a fitted gam model are provided in ?predict.gam.
_____________________________________________________________________> Simon Wood simon at stats.gla.ac.uk www.stats.gla.ac.uk/~simon/
>> Department of Statistics, University of Glasgow, Glasgow, G12 8QQ
>>> Direct telephone: (0)141 330 4530 Fax: (0)141 330 4814
_______________________________________________
R-packages mailing list
R-packages at stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/r-packages
