[Rcpp-commits] r3600 - pkg/RcppEigen/vignettes
noreply at r-forge.r-project.org
noreply at r-forge.r-project.org
Thu May 10 04:52:10 CEST 2012
Author: edd
Date: 2012-05-10 04:52:09 +0200 (Thu, 10 May 2012)
New Revision: 3600
Modified:
pkg/RcppEigen/vignettes/RcppEigen-intro-nojss.Rnw
pkg/RcppEigen/vignettes/response-to-referees.tex
Log:
added \hrule to code floats
noted that in response-to-referees draft
Modified: pkg/RcppEigen/vignettes/RcppEigen-intro-nojss.Rnw
===================================================================
--- pkg/RcppEigen/vignettes/RcppEigen-intro-nojss.Rnw 2012-05-09 12:51:42 UTC (rev 3599)
+++ pkg/RcppEigen/vignettes/RcppEigen-intro-nojss.Rnw 2012-05-10 02:52:09 UTC (rev 3600)
@@ -353,6 +353,7 @@
from its transpose and returned to \proglang{R}.
\begin{figure}[htb]
+ \hrule \smallskip
%\begin{quote}
\noindent
\ttfamily
@@ -365,6 +366,7 @@
\hlstd{}\hlkwa{return\ }\hlstd{}\hlkwd{wrap}\hlstd{}\hlsym{(}\hlstd{At}\hlsym{);}\hlstd{}\hspace*{\fill}
\normalfont
%\end{quote}
+ \hrule
\caption{\code{transCpp}: Transpose a matrix of integers.}
\label{trans}
\end{figure}
@@ -406,6 +408,7 @@
cross-product of its two arguments.
\begin{figure}[htb]
+ \hrule \smallskip
%\begin{quote}
\noindent
\ttfamily
@@ -418,6 +421,7 @@
\normalfont
\normalsize
%\end{quote}
+ \hrule
\caption{\code{prodCpp}: Product and cross-product of two matrices.}
\label{prod}
\end{figure}
@@ -473,6 +477,7 @@
multiple defaults to 1. The code in Figure~\ref{crossprod} produces
\begin{figure}[htb]
+ \hrule \smallskip
%\begin{quote}
\noindent
\ttfamily
@@ -492,6 +497,7 @@
\normalfont
\normalsize
%\end{quote}
+ \hrule
\caption{\code{crossprodCpp}: Cross-product and transposed cross-product of a single matrix.}
\label{crossprod}
\end{figure}
@@ -526,6 +532,7 @@
function, \code{AtA}, that returns the crossproduct matrix as shown in Figure~\ref{fig:incl}
\begin{figure}[htb]
+ \hrule \smallskip
%\begin{quote}
\noindent
\ttfamily
@@ -548,6 +555,7 @@
\normalfont
\normalsize
%\end{quote}
+ \hrule
\caption{The contents of the character vector, \code{incl}, that will preface \proglang{C++} code segments that follow.}
\label{fig:incl}
\end{figure}
@@ -572,6 +580,7 @@
pass a numeric matrix, $\bm A$, not an integer matrix.
\begin{figure}[htb]
+ \hrule \smallskip
%\begin{quote}
\noindent
\ttfamily
@@ -582,6 +591,7 @@
\normalfont
\normalsize
%\end{quote}
+ \hrule
\caption{\code{cholCpp}: Cholesky decomposition of a cross-product.}
\label{chol}
\end{figure}
@@ -629,6 +639,7 @@
Figure~\ref{cholDet}.
\begin{figure}[htb]
+ \hrule \smallskip
%\begin{quote}
\noindent
\ttfamily
@@ -642,6 +653,7 @@
\normalfont
\normalsize
%\end{quote}
+ \hrule
\caption{\code{cholDetCpp}: Determinant of a cross-product using
the ``LLt'' and ``LDLt'' forms of the Cholesky decomposition.}
\label{cholDet}
@@ -688,6 +700,7 @@
\label{sec:LLtLeastSquares}
\begin{figure}[tbh]
+ \hrule \smallskip
%\begin{quote}
\noindent
\ttfamily
@@ -713,6 +726,7 @@
\normalfont
\normalsize
%\end{quote}
+ \hrule
\caption{\code{lltLSCpp}: Least squares using the Cholesky decomposition.}
\label{lltLS}
\end{figure}
@@ -819,6 +833,7 @@
% const VectorXd se(QR.matrixQR().topRows(p).triangularView<Upper>()
% .solve(MatrixXd::Identity(p,p)).rowwise().norm());
\begin{figure}[htb]
+ \hrule \smallskip
%\begin{quote}
\noindent
\ttfamily
@@ -833,6 +848,7 @@
\mbox{}
\normalfont
\normalsize
+ \hrule
\caption{\code{QRLSCpp}: Least squares using the unpivoted QR decomposition.}
\label{QRLS}
%\end{quote}
@@ -937,6 +953,7 @@
\bm X^+=\bm V\bm D_1^+\bm U_1^\prime .
\end{displaymath}
\begin{figure}[htb]
+ \hrule \smallskip
%\begin{quote}
\noindent
\ttfamily
@@ -950,6 +967,7 @@
\normalfont
\normalsize
%\end{quote}
+ \hrule
\caption{\code{DplusCpp}: Create the diagonal $\bm d^+$ of the pseudo-inverse, $\bm D_1^+$, from the array of singular values, $\bm d$.}
\label{Dplus}
\end{figure}
@@ -977,6 +995,7 @@
With these definitions the code for least squares using the singular
value decomposition can be written as in Figure~\ref{SVDLS}.
\begin{figure}[htb]
+ \hrule \smallskip
%\begin{quote}
\noindent
\ttfamily
@@ -991,6 +1010,7 @@
\normalfont
\normalsize
%\end{quote}
+ \hrule
\caption{\code{SVDLSCpp}: Least squares using the SVD.}
\label{SVDLS}
\end{figure}
@@ -1044,6 +1064,7 @@
With these definitions one can adapt much of the code from the SVD
method for the eigendecomposition, as shown in Figure~\ref{SymmEigLS}.
\begin{figure}[htb]
+ \hrule \smallskip
%\begin{quote}
\noindent
\ttfamily
@@ -1057,6 +1078,7 @@
\normalfont
\normalsize
%\end{quote}
+ \hrule
\caption{\code{SymmEigLSCpp}: Least squares using the eigendecomposition.}
\label{SymmEigLS}
\end{figure}
@@ -1112,6 +1134,7 @@
\end{displaymath}
The vector $\bm Q^\prime\bm y$ is called the ``effects'' vector in \proglang{R}.
\begin{figure}[htb]
+ \hrule \smallskip
%\begin{quote}
\noindent
\ttfamily
@@ -1145,6 +1168,7 @@
\normalfont
\normalsize
%\end{quote}
+ \hrule
\caption{\code{ColPivQRLSCpp}: Least squares using the pivoted QR decomposition.}
\label{ColPivQRLS}
\end{figure}
Modified: pkg/RcppEigen/vignettes/response-to-referees.tex
===================================================================
--- pkg/RcppEigen/vignettes/response-to-referees.tex 2012-05-09 12:51:42 UTC (rev 3599)
+++ pkg/RcppEigen/vignettes/response-to-referees.tex 2012-05-10 02:52:09 UTC (rev 3600)
@@ -228,20 +228,21 @@
have point, though, inasmuch as figures 4 and 5 really flow into each
other, and have no visual separator. For the infamous book draft, I gave up
on what I had suggested to you here and just went with (the latex package)
- listings.
+ listings. [Follow-up 9 May] I just added a set of horizontal rules to each
+ of the floats. Take a look and see if you think that that makes it better.
}
\pointRaised{Comment 2}{
It might better to re-order the subsections of Section 4 as follows: \\
- * Least squares using the ``LLt'' Cholesky \\
- * Least squares using the unpivoted QR decomposition\\
- * Least squares using the SVD \\
- * Least squares using the eigendecomposition \\
- * Handling the ranking-deficient case \\
- a) using column-pivoted QR decomposition \\
- b) Using SVD \\
- * Comparative speed
+ -- Least squares using the ``LLt'' Cholesky \\
+ -- Least squares using the unpivoted QR decomposition\\
+ -- Least squares using the SVD \\
+ -- Least squares using the eigendecomposition \\
+ -- Handling the ranking-deficient case \\
+ \phantom{---} a) using column-pivoted QR decomposition \\
+ \phantom{---} b) Using SVD \\
+ -- Comparative speed
}
\reply{
But the SVD and eigendecomposition methods are set up to handle
More information about the Rcpp-commits
mailing list