From: Wolfgang Bangerth Date: Wed, 3 May 2023 21:17:30 +0000 (-0600) Subject: Let doxygen resolve a citation in step-77. X-Git-Tag: v9.5.0-rc1~257^2 X-Git-Url: https://gitweb.dealii.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=refs%2Fpull%2F15171%2Fhead;p=dealii.git Let doxygen resolve a citation in step-77. --- diff --git a/doc/doxygen/references.bib b/doc/doxygen/references.bib index cd63e23f43..2650dad946 100644 --- a/doc/doxygen/references.bib +++ b/doc/doxygen/references.bib @@ -1406,6 +1406,15 @@ URL = {https://doi.org/10.1137/0917003} } +@book{DennisSchnabel96, + author = {J. E. Dennis and R. B. Schnabel}, + title = {Numerical Methods for Unconstrained Optimization and Nonlinear Equations}, + publisher = {Society for Industrial and Applied Mathematics}, + year = {1996}, + doi = {10.1137/1.9781611971200}, + URL = {https://epubs.siam.org/doi/abs/10.1137/1.9781611971200} +} + %------------------------------------------------------------------------------- % Step 78 diff --git a/include/deal.II/sundials/kinsol.h b/include/deal.II/sundials/kinsol.h index 2bbc4465fb..22ec87d4ff 100644 --- a/include/deal.II/sundials/kinsol.h +++ b/include/deal.II/sundials/kinsol.h @@ -123,9 +123,8 @@ namespace SUNDIALS * way for furthering convergence of the nonlinear problem. This technique is * implemented in the second strategy, called Linesearch. This option employs * both the $\alpha$ and $\beta$ conditions of the Goldstein-Armijo - * linesearch algorithm given in *J. E. Dennis and R. B. Schnabel. "Numerical - * Methods for Unconstrained Optimization and Nonlinear Equations." SIAM, - * Philadelphia, 1996.*, where $\lambda$ is chosen to guarantee a sufficient + * linesearch algorithm given in @cite DennisSchnabel96 , + * where $\lambda$ is chosen to guarantee a sufficient * decrease in $F$ relative to the step length as well as a minimum step * length relative to the initial rate of decrease of $F$. One property of the * algorithm is that the full Newton step tends to be taken close to the @@ -150,9 +149,10 @@ namespace SUNDIALS * produce the new iterate. Next, the nonlinear residual function is * evaluated at the new iterate, and convergence is checked. The Picard and * fixed point methods can be significantly accelerated using Anderson's - * method. + * acceleration method. * - * The user has to provide the implementation of the following std::functions: + * The user has to provide the implementation of the following + * `std::function`s: * - reinit_vector; * and only one of * - residual; @@ -177,7 +177,7 @@ namespace SUNDIALS * scaling factors for both the solution and the residual evaluation during * convergence checks: * - get_solution_scaling; - * - get_function_scaling; + * - get_function_scaling. */ template > class KINSOL