[Mboost-commits] r797 - / pkg/mboostDevel/inst pkg/mboostDevel/tests pkg/mboostDevel/tests/Examples pkg/mboostDevel/vignettes
noreply at r-forge.r-project.org
noreply at r-forge.r-project.org
Thu Sep 25 20:34:43 CEST 2014
Author: hofner
Date: 2014-09-25 20:34:43 +0200 (Thu, 25 Sep 2014)
New Revision: 797
Added:
pkg/mboostDevel/inst/NEWS.Rd
Modified:
pkg/mboostDevel/tests/Examples/mboostDevel-Ex.Rout.save
pkg/mboostDevel/tests/birds_Biometrics.Rout.save
pkg/mboostDevel/tests/bugfixes.Rout.save
pkg/mboostDevel/tests/regtest-baselearner.Rout.save
pkg/mboostDevel/tests/regtest-blackboost.Rout.save
pkg/mboostDevel/tests/regtest-family.Rout.save
pkg/mboostDevel/tests/regtest-gamboost.Rout.save
pkg/mboostDevel/tests/regtest-glmboost.Rout.save
pkg/mboostDevel/tests/regtest-hatmatrix.Rout.save
pkg/mboostDevel/tests/regtest-inference.Rout.save
pkg/mboostDevel/vignettes/SurvivalEnsembles.Rout.save
pkg/mboostDevel/vignettes/mboost.Rout.save
pkg/mboostDevel/vignettes/mboost_illustrations.Rout.save
pkg/mboostDevel/vignettes/mboost_tutorial.Rout.save
svn_release.txt
Log:
- updated Rout.save files
- added NEWS.Rd
- updated release SOPs
Added: pkg/mboostDevel/inst/NEWS.Rd
===================================================================
--- pkg/mboostDevel/inst/NEWS.Rd (rev 0)
+++ pkg/mboostDevel/inst/NEWS.Rd 2014-09-25 18:34:43 UTC (rev 797)
@@ -0,0 +1,658 @@
+\name{NEWS}
+\title{News for Package 'mboost'}
+
+\section{Changes in mboost version 2.4-0 (2014-09-26, rZZZ)}{
+ \itemize{
+ \item added \code{confint} function to compute (bootstrap)
+ confidence intervals together with plot and print methods
+
+ \item improved \code{plot} method for varying coefficients
+ (\code{ylim} now suitable) and base-learners of factor variables.
+
+ \item tweaked \code{update} function: we now can turn the trace off
+ and specify the type of \code{risk} as well as the \code{oobweight}
+ to \code{update()}
+
+ \item \code{stabsel} has been moved to the new package \pkg{stabs}
+
+ \item changed vignette \code{mboost_tutorial} to reflect latest
+ changes in \pkg{mboost}.
+
+ \item changed plain text \code{NEWS} to \code{inst/NEWS.Rd}
+ }
+ \subsection{Bugfixes}{
+ \itemize{
+ \item \code{glmboost()$model.frame()} was broken
+ \item \code{glmboost()$update()} was broken
+ \item \code{predict()} for models with non-scalar offsets was broken
+ }
+ }
+}
+
+\section{Changes in mboost version 2.3-0 (2014-06-26, r771)}{
+ \itemize{
+ \item \code{stabsel} was recoded and now uses different terminology, much more options
+ and a better tested code base
+ \item new replacement function \code{mstop<-} as an alternative to \code{<mboost>[i]}
+ (suggested by Achim Zeileis).
+ \item \code{bmono}
+ \itemize{
+ \item new and faster algorithm to compute monotonic P-splines
+ (\code{type = "quad.prog"})
+ \item new constraints added for positive and negative spline
+ estimates
+ }
+ \item \code{bbs}
+ \itemize{
+ \item allows monotone T-splines (experimental)
+ \item new argument \code{deriv} to bbs for computing derivatives of
+ B-splines
+ }
+ \item \code{bmrf} can now also handle neighborhood matrizes as an
+ argument to \code{bnd}
+ \item added new families \code{Hurdle} and \code{Multinomial}
+ \item \code{boost_control}: added new argument \code{stopintern} for
+ internal stopping (based on oobag data) during fitting
+ \item All data sets have been moved to the new package set \pkg{TH.data}
+ }
+ \subsection{Miscellaneous}{
+ \itemize{
+ \item added new argmument \code{which} to \code{variable.names()}
+ \item added new method \code{risk} to extract risks
+ \item \code{brandom} now checks that a factor is given
+ \item speed improvements when updating a model via \code{mod[mstop]}
+ \item changed \code{\\dontrun} to \code{\\donttest}
+ \item updated references
+ }
+ }
+ \subsection{Bugfixes}{
+ \itemize{
+ \item fixed a problem with \code{extract()} of single base-learners
+ \item fixed bug in \code{AIC.mboost}: \code{df = "actset"} can
+ only be used with glmboost models
+ \item fixed package startup messages
+ \item fixed a problem in \code{mboost_fit} (when names of
+ base-learners were missing)
+ }
+ }
+}
+
+\section{Changes in mboost version 2.2-3 (2013-09-09, r733)}{
+ \itemize{
+ \item fixed bugs in survival families:
+ \itemize{
+ \item offset in all survival families was based on max(survtime) instead
+ of max(log(survtime));
+ \item offset in CoxPH can't be computed from Cox Partial LH as constants
+ are canceled out; Use fixed offset instead;
+ }
+ \item speed up checking of manual by changing some computations (e.g. reduce
+ mstop) or exclude code from checking via \\dontrun\{\}
+ \item removed dependency on ipred (replaced with TH.data)
+ \item small improvements in manual
+ }
+}
+\section{Changes in mboost version 2.2-2 (2013-02-08, r703)}{
+ \itemize{
+
+ \item bbs(..., center = "spectralDecomp") computes the spectral
+ decomposition of the penalty matrix and the penalized part of the
+ design matrix is defined by this decomposition. Experiments show
+ that \code{bols(x) + bbs(x, center = "spectralDecomp")} is a little
+ better in recovering the true underlying functions than the default
+ \code{bols(x) + bbs(x, center = TRUE)} or, equivalently,
+ \code{bols(x) + bbs(x, center = "differenceMatrix")}. For
+ \code{bbs(x, y, center = TRUE)} or \code{bmrf(x, center = TRUE)},
+ the spectral decomposition is (and was) always used.
+
+ \item fixed bug in stabsel: '...' was not passed to cvrisk and thus one could
+ not specify options for mclapply
+ \item fixed bug in brandom: now really use contrasts.arg = "contr.dummy" per
+ default.
+ \item removed tests/ folder and .Rout.save files for vignettes from the CRAN
+ release
+ \item small improvements in manual
+ }
+}
+\section{Changes in mboost version 2.2-1 (2013-01-14, r694)}{
+ \itemize{
+ \item included warnings in stabsel() for better guidiance of the
+ user:
+ \itemize{
+ \item A warning is issued if the upper bound for the FWER in stability
+ selection is greater (by a certain margin) than the specified bound.
+ \item A warning is also issued if mstop is too small to select
+ q variables.
+ }
+ \item improved output of errors and warnings in stabsel.
+ \item suppress the notes from package Matrix about method ambiguity
+ ("Note: method with signature ... chosen, ... would also be valid")
+ \item updated manual on base-learners to reflect the change in the default for
+ degrees of freedom (addtionally, all options are now discussed in a
+ separate section of the baselearner manual)
+ \item updated vignette mboost_tutorial
+ \item updated mboost_package.Rd:
+ now all important changes since mboost 2.0 are documented there
+ \item changed roles of contributers to ctb
+ \item suggested packages are now only used inside if(require(pkg)) statements
+ \item changed startup message
+ }
+}
+\section{Changes in mboost version 2.2-0 (2012-11-21, r680)}{
+ \itemize{
+ \item switch from packages `multicore' and`snow' to `parallel'
+ \item changed behavior of bols(x, intercept = FALSE) when x is a factor:
+ now the intercept is simply dropped from the design matrix;
+ coding can be specified as usually for factors.
+ \item changed default for options("mboost_dftraceS") to FALSE, i.e.,
+ degrees of freedom are now computed from smoothing parameter
+ as described in B. Hofner, T. Hothorn, T. Kneib, M. Schmid (2011).
+ \item changed computation of B-spline basis at the boundaries:
+ now also use equidistant knots in the boundaries (per default)
+ \item improved plot function when dealing with spatial plots
+ (now builds suitable grid based on the observations if no newdata is given)
+ \item increased default number of subsampling replicates in stabsel to 100
+ \item [experimental] bmono() now implements constraints at the boundaries of
+ (monotonic) P-splines
+ \item [experimental] added family Gehan() for rank-based estimation of survival models
+ in an accelerated failure time framework (contributed by Brent Johnson
+ <bajohn3 at emory.edu>)
+ }
+}
+\section{Changes in mboost version 2.1-3 (2012-09-27, r672)}{
+ \itemize{
+ \item matrices with one column are now handled as vectors in base-learners
+ \item improved manual
+ \item fixed error that occures with R (>= 2.16) due to internal changes in R
+ }
+}
+\section{Changes in mboost version 2.1-2 (2012-02-29, r647)}{
+ \itemize{
+ \item improved handling of missing values
+ (throws warnings and fixed a bug that occured for missings in the response)
+ \item improved manual for the handling of contrasts in bols
+ \item added tutorial vignette
+ \item updated references
+ }
+}
+\section{Changes in mboost version 2.1-1 (2011-11-28, r606)}{
+ \itemize{
+ \item new option "mboost_eps" for factor in Demmler-Reinsch orthogonalization
+ }
+}
+\section{Changes in mboost version 2.1-0 (2011-11-15, r601)}{
+ \itemize{
+ \subsection{Base-learners}{
+ \itemize{
+ \item added base-learners for smooth monotonic (or convex/concave)
+ functions of one or two variables (bmono())
+ \item added base-learners for radial basis functions (brad())
+ \item added base-learners for Markov random fields (bmrf())
+ \item bbs(x, cyclic = TRUE) for cyclic covariates ensures that
+ predictions at the boundaries coincide and that the resulting
+ function estimate is smoothly joined
+ \item bols(x, intercept = FALSE) only reasonable if x is centered.
+ A warning is now issued if x is not centered.
+ \item changed default for degrees of freedom in bspatial()
+ to df = 6
+ \item added checks in bbs (and brandom) to ensure that the specified
+ degrees of freedom are greater than the range of the (unpenalized)
+ null space
+ \item bolscw can be mixed with other base-learners (although
+ not yet exported and not via the formula interface)
+ \item new experimental baselearner \%O\% for smoothing
+ matrix-values responses
+ }
+ }
+ \subsection{Families}{
+ \itemize{
+ \item add Binomial(link = "probit") and general cdf's as link
+ functions (experimental)
+ \item added new families:
+ \itemize{
+ \item AUC() for AUC loss function
+ \item GammaReg() for gamma regression models
+ }
+ }
+ }
+ \subsection{Methods}{
+ \itemize{
+ \item added extract() methods for base-learners and fitted models
+ \item added residuals() function to extract residuals from the
+ model
+ \item improved predict.mboost(): added names where missing and the
+ offset as attribute where applicable.
+ \item fixed bug in predict() with glmboost.matrix(..., center = TRUE)
+ \item coef now also works with tree base-learners
+ (returns NULL in this case)
+ \item changed coef.gamboost to coef.mboost
+ \item various improvements in plot.mboost function
+ }
+ }
+
+ \subsection{Miscellaneous}{
+ \itemize{
+ \item changed default in glmboost() to center = TRUE
+ \item speed up glmboost() a little bit
+ \item changed behavior of cvrisk() if weights are used:
+ out-of-bag-risk now weighted according to "weights" as
+ specified in call to mboost
+ \item added warning if df2lambda is likely to become numerically
+ instable (i.e. in the case of large entries in the design matrix)
+ \item improved storage, speed and stability using Matrix technology for
+ bols() for factors with many levels and brandom();
+ further improvements in base-learners that are combined via \%+\%.
+ \item various improvements and fixes in manuals
+ }
+ }
+ }
+}
+\section{Changes in mboost version 2.0-12 (2011-08-22, r586)}{
+ \itemize{
+ \item minor bugfixes to make mboost work with gamboostLSS
+ \item replaced writeLines with packageStartupMessage in .onAttach()
+ \item replaced partially matched function arguments by full arguments
+ \item minor fixes in manuals
+ }
+}
+\section{Changes in mboost version 2.0-11 (2011-03-17, r561)}{
+ \itemize{
+ \item fix problem in bl_lin when using dense matrices from package "Matrix"
+ }
+}
+\section{Changes in mboost version 2.0-10 (2011-02-20, r556)}{
+ \itemize{
+ \item add rqss results for India childhood malnutrition data
+ }
+}
+\section{Changes in mboost version 2.0-9 (2010-11-19, r543)}{
+ \itemize{
+ \item add gbm to Suggests
+ }
+}
+\section{Changes in mboost version 2.0-8 (2010-11-11, r542)}{
+ \itemize{
+ \item make survival package happy again
+ }
+}
+\section{Changes in mboost version 2.0-7 (2010-09-28, r534)}{
+ \itemize{
+ \item vignette "mboost" updated
+ \item remove problem with R CMD check that occurred on some 64bit systems
+ }
+}
+\section{Changes in mboost version 2.0-6 (2010-05-22, r510)}{
+ \itemize{
+ \item no not use multicore functionality in R CMD check, really.
+ }
+}
+\section{Changes in mboost version 2.0-5 (2010-05-21, r507)}{
+ \itemize{
+ \item no not use multicore functionality in R CMD check
+ }
+}
+\section{Changes in mboost version 2.0-4 (2010-04-15, r490)}{
+ \itemize{
+ \item new vignette "mboost" describing 2.0-x series features
+ \item fixed bug in bols(): contrast.arg was ignored if not a named list
+ (which is wasn't per default)
+ \item added (missing) response functions to families Weibull(),
+ Loglog(), Lognormal() and NBinomial()
+ \item fixed bug in family CoxPH which occurred with NAs
+ \item improvements and corrections in documentation
+ }
+}
+\section{Changes in mboost version 2.0-3 (2010-03-10, r456)}{
+ \itemize{
+ \item glmboost(..., center = TRUE) now also centers columns of the
+ design matrix corresponding to contrasts of factors
+ when an intercept term is present leading to faster risk
+ minimization in these cases.
+ \item coef.glmboost: New argument `off2int = TRUE' adds the
+ offset to the intercept. In addition, the intercept
+ term is now adjusted for centered covariates.
+ \item check for infinite residuals in mboost_fit(). Especially
+ for family = Poisson(), something like boost_control(nu = 0.01)
+ fixes this problem.
+ \item "by" (in bols() and bbs()) can now handle factors with more than
+ two levels
+ \item improved plot.mboost() for varying coefficients
+ \item minor improvements in documentation
+ }
+}
+\section{Changes in mboost version 2.0-2 (2010-03-04)}{
+ \itemize{
+ \item fixed bug in helper function get_index, which caused (in some circumstances)
+ wrong handling of factors in gamboost() (spotted by
+ Juliane Schaefer <JSchaefer _at_ uhbs.ch>)
+ \item reduce memory footprint in blackboost (requires party 0.9-9993)
+ }
+}
+\section{Changes in mboost version 2.0-1 (2010-03-01)}{
+ \itemize{
+ \item fixed bug in coef( , aggregate = "cumsum"): fraction "nu" was missing
+ }
+}
+\section{Changes in mboost version 2.0-0 (2010-02-01)}{
+ \itemize{
+ \item generic implementation of component-wise functional gradient
+ boosting in `mboost_fit', specialized code for linear,
+ additive and interaction models removed
+ \item new families available for ordinal, expectile and censored regression
+ \item computations potentially based on package Matrix
+ (reduces memory usage)
+ \item various speed improvements
+ \item added interface to extract selected base-learners (selected())
+ \item added interface for parallel computations in cvrisk with
+ arbitrary packages (e.g. multicore, snow)
+ \item added "which" argument in predict and coef functions and improved
+ usability of "which" in plot-function. Users can specify "which" as
+ numeric value or as a character string
+ \item added function cv() to generate matrices for k-fold cross-validation,
+ subsampling and bootstrap
+ \item new function stabsel() for stability selection with error control
+ \item added function model.weights() to extract the weights
+ \item added interface to expand model by increasing mstop in
+ model[mstop]
+ \item alternative definition of degrees of freedom available
+ \item Interface changes:
+ \itemize{
+ \item class definition / Family() arguments changed
+ \item changed behavior of subset method (model[mstop]). Object
+ is directly altered and not duplicated
+ \item argument "center" in bols replaced with "intercept"
+ \item argument "z" in base-learners replaced with "by"
+ \item bns and bss deprecated;
+ }
+ }
+}
+\section{Changes in mboost version 1.1-4 (2009-11-18)}{
+ \itemize{
+ \item fixed bug in prediction with varying coefficients for binary
+ effect modifiers
+ }
+}
+\section{Changes in mboost version 1.1-3 (2009-09-21)}{
+ \itemize{
+ \item better x-axes in plot.cvrisk and possibility to change xlab
+ \item parallel cvrisk on Unix systems only (`multicore' isn't safe on windows)
+ \item included new penalty for ordinal predictors (in bols())
+ \item corrected bug in bspatial (centering was not used for Xna)
+ \item removed output of dfbase (which is seldom used) in gamboost
+ \item changed manual for coef.gamboost
+ \item make sure NAs are handled correctly when center = TRUE
+ in glmboost
+ }
+}
+\section{Changes in mboost version 1.1-2 (2009-07-21)}{
+ \itemize{
+ \item better weights and boundary knots handling in bspatial
+ \item cvrisk runs in parallel if package `multicore' is available
+ \item errors removed and minor improvements in the manuals
+ \item center = TRUE in glmboost did only apply to numeric (not integer)
+ predictors
+ \item for safety reasons: na.action = na.omit again
+ (causes slight changes in wpbc3 example)
+ }
+}
+\section{Changes in mboost version 1.1-1 (2009-04-21)}{
+ \itemize{
+ \item new quantile regression facilities.
+ \item fix problem with bbs baselearner and cvrisk
+ }
+}
+\section{Changes in mboost version 1.1-0 (2009-03-27)}{
+ \itemize{
+ \item bbs instead of bss is the default baselearner in gamboost
+ \item make sure bbs with weights and expanded observations
+ returns numerically the very same results
+ \item btree can now deal with multiple variables
+ \item new gMDL criterion (contributed by Zhu Wang <zhu.wang at yale.edu>)
+ \item make survival package happy again
+ }
+}
+\section{Changes in mboost version 1.0-6 (2009-01-09)}{
+ \itemize{
+ \item bols allows to specify non-default contrasts.
+ }
+}
+\section{Changes in mboost version 1.0-5 (2008-12-02)}{
+ \itemize{
+ \item remove experimental memory optimization steps
+ }
+}
+\section{Changes in mboost version 1.0-4 (2008-11-12)}{
+ \itemize{
+ \item negative gradient of GaussClass() was wrong, spotted by
+ Kao Lin <linkao at picb.ac.cn>
+ }
+}
+\section{Changes in mboost version 1.0-3 (2008-11-07)}{
+ \itemize{
+ \item Date was malformed in DESCRIPTION
+ }
+}
+\section{Changes in mboost version 1.0-2 (2008-11-05)}{
+ \itemize{
+ \item improved memory footprint in gamboost() and cvrisk()
+ \item option to suppress saving of ensembless added to
+ boost_control()
+ \item bbs(), bns(), bspatial(): default number of knots changed to
+ a fixed value (= 20)
+ \item changed default for grid (now uses all iterations) in
+ cvrisk() and changed plot.cvrisk()
+ \item bols: works now for factors and can be set-up
+ to use Ridge-estimation. Intercept can be omitted
+ now (via center = TRUE).
+ \item new btree() baselearner for gamboost() available
+ \item fix inconsistencies in regression tests
+ \item add coef.gamboost
+ \item new generic `survFit'
+ \item cosmetics for trace = TRUE
+ }
+}
+\section{Changes in mboost version 1.0-1 (2007-12-09)}{
+ \itemize{
+ \item inst/mboost_Bioinf.R was missing from mboost 1.0-0
+ }
+}
+\section{Changes in mboost version 1.0-0 (2007-11-13)}{
+ \itemize{
+ \item documentation updates
+ }
+}
+\section{Changes in mboost version 0.9-0}{
+ \itemize{
+ \item tests update and release the new version on CRAN
+ \item predict(..., allIterations = TRUE) returns the matrix
+ of predictors for all boosting iterations
+ }
+}
+\section{Changes in mboost version 0.6-2}{
+ \itemize{
+ \item move `mboost' to R-forge
+ \item improvements in `gamboost':
+ \item P-splines as base learners available
+ \item new formula interface for specifying the base learner
+ \item new plot.gamboost
+ \item add the number of selected variables as degrees of freedom
+ (as mentioned in the discussion of Hastie to Buehlmann & Hothorn)
+ \item status information during fitting is now available via
+ boost_control(trace = TRUE) but is switched off by default
+ \item acknowledge constributions by Thomas Kneib and Matthias Schmid
+ in DESCRIPTION
+ }
+}
+\section{Changes in mboost version 0.6-1}{
+ \itemize{
+ \item gamboost() now allows for user-specified base learners
+ via the formula interface
+ \item gamboost.matrix(x = x, ...) requires colnames being set
+ for `x'
+ \item na.action = na.omit fix for g\{al\}mboost()
+ }
+}
+\section{Changes in mboost version 0.5-8 (2007-05-31)}{
+ \itemize{
+ \item gamboost(..., weights = w) was broken
+ }
+}
+\section{Changes in mboost version 0.5-7 (2007-05-30)}{
+ \itemize{
+ \item extract response correctly in fitted.blackboost
+ \item hatvalues (and thus AICs) for GLMs with centering of
+ covariates may have been wrong since version 0.5-0
+ \item add paper examples to tests
+ }
+}
+\section{Changes in mboost version 0.5-6 (2007-05-07)}{
+ \itemize{
+ \item fix Rd problems
+ }
+}
+\section{Changes in mboost version 0.5-5 (2007-04-25)}{
+ \itemize{
+ \item `westbc' regenerated
+ \item LazyLoad: yes (no SaveImage: yes)
+ }
+}
+\section{Changes in mboost version 0.5-4 (2007-04-18)}{
+ \itemize{
+ \item plot() method for `glmboost' objects visualizing the
+ coefficient path (feature request by Axel Benner <benner at dkfz.de>).
+ \item predict(newdata = <matrix>) was broken for gamboost(),
+ thanks to Max Kuhn <Max.Kuhn at pfizer.com> for spotting this.
+ }
+}
+\section{Changes in mboost version 0.5-3 (2007-03-23)}{
+ \itemize{
+ \item predict() for gamboost(..., dfbase = 1) was not working correctly
+ \item small performance and memory improvements for glmboost()
+ }
+}
+\section{Changes in mboost version 0.5-2 (2007-02-28)}{
+ \itemize{
+ \item some performance improvements for `glmboost()'
+ \item blackboost() is now generic with formula and x, y interface
+ \item plot() method for cvrisk() and AIC() output now allows for ylim
+ specification without troubles
+ }
+}
+\section{Changes in mboost version 0.5-1 (2007-02-02)}{
+ \itemize{
+ \item depends party 0.9-9
+ }
+}
+\section{Changes in mboost version 0.5-0 (2007-01-30)}{
+ \itemize{
+ \item new `baselearner' argument to `gamboost' allowing to
+ specify difference component-wise base-learners to
+ be used. Currently implemented: "ssp" for smoothing splines
+ (default), "bsp" for B-splines and "ols" for linear models.
+ The latter two haven't been tested yet.
+ \item The `dfbase' arguments now applies to each covariate and
+ no longer to each column of the design matrix.
+ \item cvrisk() for blackboost() was broken, totally :-(
+ \item centered covariates were returned by glmboost() and gamboost()
+ \item Poisson() used an incorrect offset
+ \item check for y being positive counts when family = "Poisson()"[B
+ \item checks for Poisson() logLik() and AIC() methods
+ \item fire a warning when all u > 0 or u < 0
+ \item update vignette `mboost_illustrations'
+ }
+}
+\section{Changes in mboost version 0.4-17 (2007-01-15)}{
+ \itemize{
+ \item fix problem with `dfbase' in `gamboost', spotted by
+ Karin Eckel <Karin.Eckel at imbe.imed.uni-erlangen.de>
+ }
+}
+\section{Changes in mboost version 0.4-16 (2007-01-12)}{
+ \itemize{
+ \item work around stats4:::AIC
+ }
+}
+\section{Changes in mboost version 0.4-15 (2006-12-06)}{
+ \itemize{
+ \item fix plot problems in plot.cvrisk
+ \item allow for centering of the numerical covariates in glmboost and
+ gamboost
+ }
+}
+\section{Changes in mboost version 0.4-14 (2006-10-27)}{
+ \itemize{
+ \item AIC(..., "classical") is now faster for non-gaussian families
+ }
+}
+\section{Changes in mboost version 0.4-13 (2006-10-04)}{
+ \itemize{
+ \item predict(..., newdata) can take a matrix now
+ }
+}
+\section{Changes in mboost version 0.4-12 (2006-09-13)}{
+ \itemize{
+ \item predict(<blackboost-object>, type = "response") did not return
+ factors when the response was actually a factor
+ \item report offset in print methods
+ \item add offset attribute to coef.glmboost
+ }
+}
+\section{Changes in mboost version 0.4-11 (2006-09-07)}{
+ \itemize{
+ \item add `contrasts.arg' argument to `glmboost.formula'
+ \item more meaningful default for `grid' in `cvrisk'
+ \item R-2.4.0 fixes
+ }
+}
+\section{Changes in mboost version 0.4-10 (2006-08-30)}{
+ \itemize{
+ \item add checks for CoxPH (against coefficients and logLik of coxph)
+ \item add weights to CoxPH
+ \item the ngradient function in Family objects needs to implement
+ arguments (y, f, w), not just (y, f)
+ \item check for meaningful class of the response for some families
+ }
+}
+\section{Changes in mboost version 0.4-9 (2006-07-17)}{
+ \itemize{
+ \item some small speed improvements in `gamboost'
+ \item handle factors in `gamboost' properly (via a linear model)
+ \item the dfbase argument can take a vector now (in `gamboost')
+ \item update and improve entries in DESCRIPTION
+ \item documentation updates
+ }
+}
+\section{Changes in mboost version 0.4-8 (2006-07-05)}{
+ \itemize{
+ \item Huber() is `Huber Error', not `Huber Absolute Error'
+ \item added `CoxPH' family object for fitting Cox models
+ \item remove inst/LaTeX
+ \item use NROW / NCOL more often (now that `y' may be a `Surv' object)
+ \item implement `cvrisk', a general cross-validation function for the
+ empirical risk and a corresponding plot method
+ \item unify risk computations in all three fitting functions
+ \item unify names for `gb' objects
+ \item allow for out-of-bag risk computations
+ \item some cosmetics
+ \item update keywords in Rd-files
+ \item risk was always 0 in Huber()@risk when d was choosen adaptively
+ \item pData(westbc)$nodal.y has levels `negative' and `positive'
+ (lymph node status)
+ }
+}
+\section{Changes in mboost version 0.4-7 (2006-06-19)}{
+ \itemize{
+ \item add src/Makevars (required for Windows builds)
+ \item make sure objects that are modified at C-level get _copied_ in
+ `blackboost'
+ }
+}
+\section{Changes in mboost version 0.4-6 (2006-06-14)}{
+ \itemize{
+ \item some minor `codetools' fixes: removed unused variables
+ and an out-dated function
+ \item add `codetools' checks to regression tests
+ \item fix xlab in plot.gbAIC
+ `mboost' version 0.4-5 published on CRAN 2006-06-13
+ }
+}
Modified: pkg/mboostDevel/tests/Examples/mboostDevel-Ex.Rout.save
===================================================================
--- pkg/mboostDevel/tests/Examples/mboostDevel-Ex.Rout.save 2014-09-25 15:33:47 UTC (rev 796)
+++ pkg/mboostDevel/tests/Examples/mboostDevel-Ex.Rout.save 2014-09-25 18:34:43 UTC (rev 797)
@@ -1,5 +1,5 @@
-R version 3.1.0 (2014-04-10) -- "Spring Dance"
+R version 3.1.1 (2014-07-10) -- "Sock it to Me"
Copyright (C) 2014 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)
@@ -34,7 +34,7 @@
> ### * </HEADER>
> library('mboostDevel')
Loading required package: parallel
-This is mboostDevel 2.3-0. See ‘package?mboostDevel’ and the NEWS file
+This is mboostDevel 2.4-0. See ‘package?mboostDevel’ and the NEWS file
for a complete list of changes.
>
@@ -947,6 +947,135 @@
> base::assign(".dptime", (proc.time() - get(".ptime", pos = "CheckExEnv")), pos = "CheckExEnv")
> base::cat("boost_family-class", base::get(".format_ptime", pos = 'CheckExEnv')(get(".dptime", pos = "CheckExEnv")), "\n", file=base::get(".ExTimings", pos = 'CheckExEnv'), append=TRUE, sep="\t")
> cleanEx()
+> nameEx("confint")
+> ### * confint
+>
+> flush(stderr()); flush(stdout())
+>
+> base::assign(".ptime", proc.time(), pos = "CheckExEnv")
+> ### Name: confint.mboost
+> ### Title: Pointwise Bootstrap Confidence Intervals
+> ### Aliases: confint.mboost confint.glmboost plot.mboost.ci lines.mboost.ci
+> ### print.glmboost.ci
+> ### Keywords: methods
+>
+> ### ** Examples
+>
+> ### a simple linear example
+> set.seed(1907)
+> data <- data.frame(x1 = rnorm(100), x2 = rnorm(100),
++ z = factor(sample(1:3, 100, replace = TRUE)))
+> data$y <- rnorm(100, mean = data$x1 - data$x2 - 1 * (data$z == 2) +
++ 1 * (data$z == 3), sd = 0.1)
+> linmod <- glmboost(y ~ x1 + x2 + z, data = data,
++ control = boost_control(mstop = 200))
+>
+> ## compute confidence interval from 10 samples. Usually one should use
+> ## at least 1000 samples.
+> CI <- confint(linmod, B = 10, level = 0.9)
+Warning in cvrisk.mboost(mod, folds = cv(model.weights(mod), B = B.mstop), :
+ zero weights
+Warning in cvrisk.mboost(mod, folds = cv(model.weights(mod), B = B.mstop), :
+ zero weights
+Warning in cvrisk.mboost(mod, folds = cv(model.weights(mod), B = B.mstop), :
+ zero weights
+Warning in cvrisk.mboost(mod, folds = cv(model.weights(mod), B = B.mstop), :
+ zero weights
+Warning in cvrisk.mboost(mod, folds = cv(model.weights(mod), B = B.mstop), :
+ zero weights
+Warning in cvrisk.mboost(mod, folds = cv(model.weights(mod), B = B.mstop), :
+ zero weights
+Warning in cvrisk.mboost(mod, folds = cv(model.weights(mod), B = B.mstop), :
+ zero weights
+Warning in cvrisk.mboost(mod, folds = cv(model.weights(mod), B = B.mstop), :
+ zero weights
+Warning in cvrisk.mboost(mod, folds = cv(model.weights(mod), B = B.mstop), :
+ zero weights
+Warning in cvrisk.mboost(mod, folds = cv(model.weights(mod), B = B.mstop), :
+ zero weights
+> CI
+ Bootstrap Confidence Intervals
+ 5% 95%
+(Intercept) -0.01975195 0.02598639
+x1 0.98986917 1.01853830
+x2 -1.00530226 -0.98158388
+z2 -1.04350674 -0.99290512
+z3 0.98234788 1.06652604
+>
+> ## to compute a confidence interval for another level simply change the
+> ## level in the print function:
+> print(CI, level = 0.8)
+ Bootstrap Confidence Intervals
+ 10% 90%
+(Intercept) -0.01105303 0.01607163
+x1 0.99577904 1.01340240
+x2 -1.00234059 -0.98355672
+z2 -1.03767821 -0.99552711
+z3 0.98851355 1.04334685
+> ## or print a subset (with point estimates):
+> print(CI, level = 0.8, pe = TRUE, which = "z")
+ Bootstrap Confidence Intervals
+ beta 10% 90%
+z2 -1.016220 -1.0376782 -0.9955271
+z3 1.024731 0.9885135 1.0433468
+>
+>
+> ### a simple smooth example
+> set.seed(1907)
+> data <- data.frame(x1 = rnorm(100), x2 = rnorm(100))
+> data$y <- rnorm(100, mean = data$x1^2 - sin(data$x2), sd = 0.1)
+> gam <- gamboost(y ~ x1 + x2, data = data,
++ control = boost_control(mstop = 200))
+>
+> ## compute confidence interval from 10 samples. Usually one should use
+> ## at least 1000 samples.
+> CI_gam <- confint(gam, B = 10, level = 0.9)
+Start computing bootstrap confidence intervals...
+
+B = 1Warning in cvrisk.mboost(object = list(baselearner = list(`bbs(x1, df = dfbase)` = list( :
+ zero weights
+
+B = 2Warning in cvrisk.mboost(object = list(baselearner = list(`bbs(x1, df = dfbase)` = list( :
+ zero weights
+
+B = 3Warning in cvrisk.mboost(object = list(baselearner = list(`bbs(x1, df = dfbase)` = list( :
+ zero weights
+
+B = 4Warning in cvrisk.mboost(object = list(baselearner = list(`bbs(x1, df = dfbase)` = list( :
+ zero weights
+
+B = 5Warning in cvrisk.mboost(object = list(baselearner = list(`bbs(x1, df = dfbase)` = list( :
+ zero weights
+
+B = 6Warning in cvrisk.mboost(object = list(baselearner = list(`bbs(x1, df = dfbase)` = list( :
+ zero weights
+
+B = 7Warning in cvrisk.mboost(object = list(baselearner = list(`bbs(x1, df = dfbase)` = list( :
+ zero weights
+
+B = 8Warning in cvrisk.mboost(object = list(baselearner = list(`bbs(x1, df = dfbase)` = list( :
+ zero weights
+
+B = 9Warning in cvrisk.mboost(object = list(baselearner = list(`bbs(x1, df = dfbase)` = list( :
+ zero weights
+
+B = 10Warning in cvrisk.mboost(object = list(baselearner = list(`bbs(x1, df = dfbase)` = list( :
+ zero weights
+
+>
+> par(mfrow = c(1, 2))
+> plot(CI_gam, which = 1)
+> plot(CI_gam, which = 2)
+> ## to compute a confidence interval for another level simply change the
+> ## level in the plot or lines function:
+> lines(CI_gam, which = 2, level = 0.8)
+>
+>
+>
+> base::assign(".dptime", (proc.time() - get(".ptime", pos = "CheckExEnv")), pos = "CheckExEnv")
+> base::cat("confint", base::get(".format_ptime", pos = 'CheckExEnv')(get(".dptime", pos = "CheckExEnv")), "\n", file=base::get(".ExTimings", pos = 'CheckExEnv'), append=TRUE, sep="\t")
+> graphics::par(get("par.postscript", pos = 'CheckExEnv'))
+> cleanEx()
> nameEx("cvrisk")
> ### * cvrisk
>
@@ -1414,15 +1543,44 @@
+ response = DEXfat))
> plot(mod, ask = FALSE, main = "base-learner")
>
+> ### use stability selection for base-learner selection
+> library("stabs")
+Note: the specification for class “boost_family” in package ‘mboost’ seems equivalent to one from package ‘mboostDevel’: not turning on duplicate class definitions for this class.
+Note: the specification for class “boost_family_glm” in package ‘mboost’ seems equivalent to one from package ‘mboostDevel’: not turning on duplicate class definitions for this class.
+> ## fit a linear model
[TRUNCATED]
To get the complete diff run:
svnlook diff /svnroot/mboost -r 797
More information about the Mboost-commits
mailing list