* Minor change
** Major change
2.7-1
* More rigorous initialization at C level to prevent possible
memory access problems
* Fixed predict() for types 'vars', 'nvars', and 'ngroups' with
multivariate outcomes
* As a consequence, summary(cvfit) now works for multivariate
outcomes (thank you to Cajo ter Braak for pointing out that
that this was broken)
2.7-0
* Internal restructuring: .Call now used instead of .C
* Added support for Poisson regression
* Fixed bug in cv.grpreg when attempting to use leave-one-out
cross-validation (thank you to Cajo ter Braak for pointing this
out)
2.6-0
** Various internal changes to make the package more efficient for
large data sets
2.5-0
** Added group exponential lasso method.
* Added gmax option
* Added nvars and ngroups option to predict
* Modified appearance of summary.cv.grpreg display.
2.4-0
** Added options in plot.cv.grpreg to plot estimates of r-squared,
signal-to-noise ratio, scale parameter, and prediction error in
addition to cross-validation error (deviance)
** grpreg and cv.grpreg now allow matrix y to facilitation group
penalized methods for seemingly unrelated regressions/multitask
learning. This is something of a 'beta' release at this point,
and will be developed and refined further in future releases.
* Added summary method for cv.grpreg objects
* Added coef and predict methods for cv.grpreg objects
* Fixed bug in predict type='coefficients' when 'lambda'
argument specified.
* Brought gBridge up to date so that it now handles constant
columns, etc. (see 2.2-0)
* Fixed bug in cv.grpreg with user-defined lambda values.
2.3-0
* Switched to SVD-based orthogonalization to allow for linear
dependency within groups
2.2-1
* Fixed compilation error for 32-bit Windows
* Fixed bug in calculation of binomial deviance when fitted
probabilities close to 0 or 1 arise
2.2-0
* cv.grpreg: Now returns full data fit as well as CV errors
* The above thereby allows cv.grpreg to handle constant columns,
and fixes some bugs
* select: Now allows ... options to be passed to logLik
* plot: Added option to plot norm of each group, rather than
individual coefficients
* predict: "vars", "groups", and "norm" options added
* logLik: fixed bug -- no longer calculates (meaningless)
log-likelihoods for saturated models (thank you to Xiaowei Ren
for pointing this out)
* fixed bug for returning group when some groups were eliminated
due to constant columns
2.1-0
* Fixed bug involving orthogonalization with unpenalized groups
* grpreg can now handle constant columns (they produce beta=0)
* Internal restructuring of C code
2.0-0
** New algorithm for group lasso
** (L2) Group MCP, group SCAD methods added
** Added cv.grpreg to facilitate cross-validation
* Extensive internal refactoring of code
* Added dfmax option
* Added group.multiplier option
* Allows specification of unpenalized groups
* standardize and orthogonalize functions added
* gBridge now divorced from grpreg and given separate function
* Much more extensive and reproducible code testing
1.2-0
* DESCRIPTION: Fixed contact info
* CITATION: Updated citation
* grpreg: Removed 'monitor' option
* grpreg: Changed 'n.lambda' to 'nlambda'
* grpreg: Changed 'a' to 'gamma' for MCP tuning parameter
* grpreg: Changed 'lambda2' to 'alpha'
* grpreg: Added 'loss' to value returned
** grpreg.c: Fixed bug in calculation of df for gLasso
* logLik: Added logLik method
** select: Syntax modified (no longer requires X, y to be passed)
* criteria: Obsolete, removed
** plot.grpreg: Made plotting function more flexible