]> https://gitweb.dealii.org/ - dealii.git/commitdiff
Fix documentation of Chebyshev iteration 8379/head
authorMartin Kronbichler <kronbichler@lnm.mw.tum.de>
Thu, 11 Jul 2019 20:14:11 +0000 (22:14 +0200)
committerMartin Kronbichler <kronbichler@lnm.mw.tum.de>
Thu, 11 Jul 2019 20:14:11 +0000 (22:14 +0200)
include/deal.II/lac/precondition.h

index b840a8087d3f80c91042ced3dacddf64a9652770..a0809713552e47608ec72aea488b354604529bfe 100644 (file)
@@ -879,7 +879,7 @@ private:
  *  x^{n+1} = x^{n} + \rho_n \rho_{n-1} (x^{n} - x^{n-1}) +
  *     \frac{\rho_n}{\lambda_{\max{}}-\lambda_{\min{}}} P^{-1} (b-Ax^n).
  * @f]
- * where the parameter $\rho_0$ is set to $\rho_0 =
+ * where the parameter $\rho_0$ is set to $\rho_0 = 2
  * \frac{\lambda_{\max{}}-\lambda_{\min{}}}{\lambda_{\max{}}+\lambda_{\min{}}}$
  * for the maximal eigenvalue $\lambda_{\max{}}$ and updated via $\rho_n =
  * \left(2\frac{\lambda_{\max{}}+\lambda_{\min{}}}
@@ -910,9 +910,9 @@ private:
  *
  * The Chebyshev method relies on an estimate of the eigenvalues of the matrix
  * which are computed during the first invocation of vmult(). The algorithm
- * invokes a conjugate gradient solver so symmetry and positive definiteness
- * of the (preconditioned) matrix system are requirements. The eigenvalue
- * algorithm can be controlled by
+ * invokes a conjugate gradient solver (i.e., Lanczos iteration) so symmetry
+ * and positive definiteness of the (preconditioned) matrix system are
+ * requirements. The eigenvalue algorithm can be controlled by
  * PreconditionChebyshev::AdditionalData::eig_cg_n_iterations specifying how
  * many iterations should be performed. The iterations are started from an
  * initial vector that depends on the vector type. For the classes

In the beginning the Universe was created. This has made a lot of people very angry and has been widely regarded as a bad move.

Douglas Adams


Typeset in Trocchi and Trocchi Bold Sans Serif.