News for Package "RcppNumerical"

Changes in RcppNumerical version 0.7-0

NEW FEATURES

• Added the optim_lbfgsb() function for box-constrained optimization.

• Multi-dimensional integration now supports infinite limits in each dimension.

• Updated LBFGS++ to v0.4.0.

BUG FIXES

• Replaced Rf_error() with Rcpp::stop() for better error handling, thanks to @Enchufa2 (#14).

Changes in RcppNumerical version 0.6-0

NEW FEATURES

• Updated LBFGS++ to v0.3.0.

Changes in RcppNumerical version 0.5-0

NEW FEATURES

• Updated LBFGS++ to v0.2.0.

BUG FIXES

• Removed the C function sprintf() calls according to CRAN's policy.

Changes in RcppNumerical version 0.4-0

NEW FEATURES

• Added one-dimensional numerical integration on infinite intervals, contributed by @rstub (#11).

Changes in RcppNumerical version 0.3-3

BUG FIXES

• Updated LBFGS++ to the newest version.

• Fixed the VignetteBuilder entry in the DESCRIPTION file according to CRAN's policy.

Changes in RcppNumerical version 0.3-2

BUG FIXES

• Fixed a convergence test bug in LBFGS++.

• Registered native routines per CRAN's policy.

Changes in RcppNumerical version 0.3-1

BUG FIXES

• Fixed a bug of name hiding in C++.

Changes in RcppNumerical version 0.3-0

NEW FEATURES

• Now using the LBFGS++ library for L-BFGS optimization.

• Added a vignette.

BUG FIXES

• Fixed an index-out-of-bound bug in one dimensional integration (#1), thanks to Greg Finak.

• Added virtual destructors for classes Func, MFunc, and MFuncGrad.

Changes in RcppNumerical version 0.2-0

NEW FEATURES

• Added the Cuba library for multi-dimensional integration.

Changes in RcppNumerical version 0.1-0

NEW FEATURES

• Initial release.

• Provides C++ interface for one-dimensional numerical integration and L-BFGS optimization.

• R function fastLR() for fast logistic regression using L-BFGS.