Merge lp:~meg-simula/fenics-book/index-fixes into lp:fenics-book

[Marie Rognes]

=== modified file 'tex/3.tex'
--- tex/3.tex	2011-10-17 21:29:59 +0000
+++ tex/3.tex	2011-10-17 22:23:24 +0000
@@ -48,6 +48,7 @@
 determining a polynomial based on interpolation of function values and
 derivatives at a set of points was also discussed in
 \citet{BrambleZlamal1970}, although the term unisolvence was not used.
+\index{unisolvence}
 
 For any finite element, one may define a local basis for
 $\CiarletSpace$ that is dual to the degrees of freedom. Such a basis
@@ -55,6 +56,7 @@
 \phi^T_j ) = \delta_{ij} $ for $1 \leqslant i,j \leqslant n $ and is called
 the \emph{nodal basis}. It is typically this basis that is used in
 finite element computations.
+\index{nodal basis}
 
 Also associated with a finite element is a \emph{local interpolation
   operator}, sometimes called a \emph{nodal interpolant}. Given some
@@ -89,6 +91,8 @@
 $\hat{\ell} \in \hat{\CiarletSpace}'$, its \emph{pushforward} acts on
 a function in $v \in \CiarletSpace$ by $\mathcal{F}_*(\hat{\ell})(v) =
 \hat{\ell}(\mathcal{F}^*(v))$.
+\index{affine equivalence}
+
 %%
 \begin{definition}[Affine equivalence]
   Let $(\hat{T}, \hat{\CiarletSpace}, \hat{\mathcal{L}})$ and
@@ -141,6 +145,7 @@
 %------------------------------------------------------------------------------
 \section{Notation}
 
+\index{degrees of freedom}
 \begin{itemize}
 \item
   The space of polynomials of degree up to and including $q$ on a
@@ -246,6 +251,7 @@
 
 \vspace*{7pt}
 \subsection{The Lagrange element}
+\index{Finite element!Lagrange}
 
 The best-known and most widely used finite element is the $\Poly{1}$
 Lagrange element. This lowest-degree triangle is sometimes called the
@@ -335,6 +341,7 @@
 by using a Lagrange element for each component.
 
 \subsection{The Crouzeix--Raviart element}
+\index{Finite element!Crouzeix--Raviart}
 
 {The Crouzeix--Raviart element was introduced in
 \citet{CrouzeixRaviart1973} as a technique for solving the stationary
@@ -442,6 +449,7 @@
 
 \vspace*{-6pt}\subsection{The Raviart--Thomas element}
 \label{sec:raviartthomas}
+\index{Finite element!Raviart--Thomas}
 
 The Raviart--Thomas element was introduced
 by \citet{RaviartThomas1977}. It was the first element to discretize
@@ -556,6 +564,7 @@
 
 %------------------------------------------------------------------------------
 \subsection{The Brezzi--Douglas--Marini element}
+\index{Finite element!Brezzi--Douglas--Marini}
 
 The Brezzi--Douglas--Marini element was introduced by Brezzi,
 Douglas and Marini in two dimensions (for triangles) in
@@ -622,6 +631,7 @@
 in \citet{BrezziDouglasMarini1985a}.
 
 \subsection{The Mardal-Tai-Winther element}
+\index{Finite element!Mardal--Tai--Winther}
 
 The Mardal--Tai--Winther element was introduced
 in \citet{MardalTaiWinther2002} as a finite element suitable for the
@@ -678,6 +688,7 @@
 \end{equation}
 
 \subsection{The Arnold--Winther element}
+\index{Finite element!Arnold--Winther}
 
 The Arnold--Winther element was introduced
 by \citet{ArnoldWinther2002}. This paper presented the first stable
@@ -785,6 +796,7 @@
 
 %------------------------------------------------------------------------------
 \subsection{The \nedelec{} $\Hcurl$ element of the first kind}
+\index{Finite element!N\'ed\'elec}
 \label{sec:nedelec:first}
 
 \begin{definition}[\nedelec{} $\Hcurl$ element of the first kind]
@@ -869,6 +881,7 @@
 
 \subsection{The $\Hcurl$ \nedelec{} element of the second kind}
 \label{sec:nedelec:second}
+\index{Finite element!N\'ed\'elec}
 
 \begin{definition}[\nedelec{} $\Hcurl$ element of the second kind]
   The \nedelec{} element of the second kind ($\mathrm{NED}^2_q$) is
@@ -963,6 +976,7 @@
 an increased opportunity for parallelism and $hp$-adaptivity.
 
 \subsection{Discontinuous Lagrange}
+\index{Finite element!Discontinuous Lagrange}
 
 \begin{definition}[Discontinuous Lagrange element]
   The discontinuous Lagrange element ($\mathrm{DG}_q$)
@@ -1052,6 +1066,7 @@
 
 
 \subsection{The Argyris element}
+\index{Finite element!Argyris}
 
 The Argyris element \citep{ArgyrisFriedScharpf1968,Ciarlet2002} is
 based on the space $\Poly{5}(T)$ of quintic polynomials over some triangle
@@ -1113,6 +1128,7 @@
 continuity at the vertices \citep{SolinSegethDolevzel2004}.
 
 \subsection{The Hermite element}
+\index{Finite element!Hermite}
 
 \looseness-1{}The Hermite element generalizes the classic cubic Hermite
 interpolating polynomials on the line segment
@@ -1182,6 +1198,7 @@
 
 
 \subsection{The Morley element}
+\index{Finite element!Morley}
 
 The Morley triangle defined in \citet{Morley1968} is a simple
 $H^2$-nonconforming quadratic element that is used in fourth-degree
@@ -1264,6 +1281,8 @@
 Raviart--Thomas element enriched by the curl of the cubic bubble
 element \citep{ArnoldBrezziDouglas1984}.
 
+\index{Finite element!Bubble elements}
+
 \begin{definition}[Bubble element]
   The bubble element ($B_q$) is defined for $q \geqslant (d+1)$ by
   \begin{align}
@@ -1296,6 +1315,8 @@
 
 %------------------------------------------------------------------------------
 \section{Finite element exterior calculus}
+\index{finite element exterior calculus}
+
 It has recently been demonstrated that many of the finite elements
 that have been discovered or invented over the years can be formulated
 and analyzed in a common unifying framework as special cases of a more

=== modified file 'tex/31.tex'
--- tex/31.tex	2011-10-17 18:35:28 +0000
+++ tex/31.tex	2011-10-17 22:23:24 +0000
@@ -4,6 +4,8 @@
               {Lyudmyla Vynnytska,  Stuart R.~Clark and Marie E.~Rognes}
               {vynnytska}
 
+\index{mantle convection}
+
 In this chapter, we model dynamic convection processes in the Earth's
 mantle; linking the geodynamical equations, numerical implementation and
 Python code tightly together. The convection of material is generated by
@@ -124,6 +126,7 @@
 are nondimensional; scaling and physical constants are presented in
 Section~\ref{vynnytska:sec:results}.
 
+\index{Rayleigh number}
 The Rayleigh numbers measure the relative importance of buoyancy to
 thermal and viscous dissipation given by
 thermal diffusivity ($k_{th}$) and the reference viscosity ($\eta_0$),
@@ -142,6 +145,8 @@
 within the ranges for the Earth \citep{MontagueKelloggManga1998} and
 such that fluid convection dominates.
 
+\index{advection-diffusion equation}
+
 The mantle flow induces transport of the composition $\phi$. This
 transport is governed by the equation
 \begin{equation}
@@ -252,6 +257,7 @@
 
 
 \subsection{Mixed finite element formulation of the Stokes equations}
+\index{Stokes equations!mixed finite element discretization}
 
 Let $\mathcal{T}_h = \{T\}$ be a mesh\footnote{Note that $T$ is used
 to denote an element (cell) in a mesh in this section, in addition to
@@ -287,6 +293,8 @@
 \enlargethispage{12pt}
 
 \vspace*{-6pt}\subsection{Discontinuous Galerkin formulation of advection--diffusion equations}
+\index{discontinuous Galerkin}
+\index{advection-diffusion equation!discontinuous Galerkin discretization}
 
 The energy and transport equations~\eqref{vynnytska:eq:energy}
 and~\eqref{vynnytska:eq:transdif} have the same structure from the
@@ -376,6 +384,8 @@
 
 \subsection{A decoupling predictor-corrector scheme}
 
+\index{predictor-corrector scheme}
+
 Instead of solving the fully coupled nonlinear system of equations
 defined
 by~\eqref{vynnytska:eq:momentum}~--~\eqref{vynnytska:eq:incompress},
@@ -468,6 +478,8 @@
 where $h_{\min}$ is the minimum cell size of the mesh and
 $C_{\mathrm{CFL}}$ is a chosen positive number.
 
+\index{filtering}
+
 The basic idea of the filtering algorithm is to ensure that $\phi$
 remains within the bounds $0 \leqslant \phi \leqslant 1$, and to minimize
 dispersion error.  We refer the reader to~\citet{LenardicKaula1993}
@@ -528,6 +540,7 @@
   with the time required for the assembly and solution of the linear
   systems.
 \item
+  \index{Stokes equations!iterative linear solvers}
   The linear systems resulting from the equations for the composition
   and the temperature are positive definite but not symmetric. Hence,
   these are solved iteratively using a standard generalized minimal
@@ -922,6 +935,8 @@
 captions to Figures~\ref{vynnytska:fig:BG} and~\ref{vynnytska:fig:HM}
 for details.
 
+\index{root-mean-square velocity}
+
 \begin{table}
   \begin{tabular}{lccc}
     \toprule

=== modified file 'tex/36.tex'
--- tex/36.tex	2011-10-17 16:09:26 +0000
+++ tex/36.tex	2011-10-17 22:23:24 +0000
@@ -1,4 +1,4 @@
-\begingroup
+b\begingroup
 \fenicschapter{Automated testing of saddle point stability conditions}
               {Automated testing of saddle point stability \\ conditions}
               {Automated testing of saddle point stability conditions}
@@ -15,6 +15,9 @@
 \newcommand{\rognesascot}{ASCoT}
 \newcommand{\rognespython}{Python}
 
+\index{mixed finite element methods}
+\index{discrete stability}
+
 Over the last five decades, there has been a substantial body of
 research on the theory of mixed finite element methods.  Mixed finite
 element methods are finite element methods where two or more finite
@@ -303,6 +306,8 @@
 
 \subsection{Stability conditions for saddle point problems}
 
+\index{saddle point problems}
+
 We now turn to consider the special case of abstract saddle point
 problems. In this case, the stability
 condition~\eqref{rognes:eq:infsup:discrete} can be rephrased in an
@@ -823,6 +828,8 @@
 
 \subsection{Mixed Laplacian}
 
+\index{Poisson equation!mixed finite element discretization}
+
 We can now return to the mixed Laplacian example described in
 Section~\ref{rognes:sec:motivation} and inspect the Brezzi stability
 properties of the element spaces involved, namely
@@ -967,6 +974,8 @@
 
 \subsection{Stokes}
 
+\index{Stokes equations!mixed finite element discretization}
+
 The Stokes equations is another classical and highly relevant saddle
 point problem. For simplicity, we here consider the following discrete
 formulation: find the velocity $u_h \in V_h$, and the pressure $p_h