Algorithm with beamer package causing nightmares

Question

I have an algorithm which works well with \documentclass[twocolumn]{svjour3}. Now I have to present the paper and I would like one slide to be that algorithm, so I do:

\documentclass{beamer}
\usepackage{graphicx}

% definitions 
\newcommand{\kd}{\emph{k}-d\space}
\newcommand{\kgeraf}{\kd GeRaF}
\newcommand{\geraf}{{\tt GeRaF}\xspace}
% \newcommand{\att}[1]{\alert{#1}}
\newcommand{\att}[1]{{#1}}

% shortcuts 
\newcommand{\tab}{\hspace*{2em}}
\newcommand{\head}[1]{{\medskip\noindent\emph{#1}}}
\newtheorem{defn}{Definition}
\usepackage[top=2.5cm,bottom=2.3cm,hmargin=1.5cm]{geometry} % it was hmargin=2.5cm
\usepackage{amssymb,graphicx,url}
\setcounter{tocdepth}{3}
\usepackage[english]{babel}
\usepackage{multirow}
\usepackage[dvipsnames,svgnames]{xcolor}
\usepackage[ruled,vlined,linesnumbered]{algorithm2e}
\usepackage{xspace}
\usepackage{pgfplots}
%\pgfplotsset{compat=1.90} % ize got latex error here
\usepackage{pgfplotstable}
\def\real{{\mathbb R}}

%--------------------------------------------------------------------------------
% Add a period to the end of an abbreviation unless there's one
% already, then \xspace.
\makeatletter
\DeclareRobustCommand\onedot{\futurelet\@let@token\@onedot}
\def\@onedot{\ifx\@let@token.\else.\null\fi\xspace}
%
\def\eg{\emph{e.g}\onedot} \def\Eg{\emph{E.g}\onedot}
\def\ie{\emph{i.e}\onedot} \def\Ie{\emph{I.e}\onedot}
\def\cf{\emph{c.f}\onedot} \def\Cf{\emph{C.f}\onedot}
\def\etc{\emph{etc}\onedot} %\def\vs{\emph{vs}\onedot}
\def\wrt{w.r.t\onedot} \def\dof{d.o.f\onedot}
\def\etal{\emph{et al}\onedot}
\makeatother

%--------------------------------------------------------------------
\begin{document}
\begin{algorithm}[t]
\SetFuncSty{textsc}
\SetDataSty{emph}
\newcommand{\expr}[1]{$\langle${#1}$\rangle$}
\newcommand{\commentsty}[1]{{\color{DarkGreen}#1}}
\SetCommentSty{commentsty}
\SetKwComment{Comment}{$\triangleright$ }{}
\SetKwBlock{Block}{function}{}
\SetKwInOut{Input}{input}
\SetKwInOut{Output}{output}
\SetKw{Return}{return}
\SetKwData{Node}{node}
\SetKwData{Leaf}{leaf}
\SetKwFunction{Build}{build}
\SetKwFunction{Variance}{variance}
\SetKwFunction{Median}{median}
\SetKwFunction{Split}{split}

\Input
{pointset $X$, \#trees $m$, \#split-dimensions $t$, max \#points per leaf $p$
}
\Output{randomized \kd forest $F$}
\Begin
{
    $V \gets$ \expr{\Variance of $X$ in every dimension} \\
    $D \gets$ \expr{$t$ dimensions of maximum variance $V$} \\
    $F \gets \emptyset$  \Comment*[f]{forest} \\
    \For{$i \gets 1$ \KwTo $m$}
    {
        $f \gets$ \expr{random transformation}  \Comment*[f]{isometry, shuffling} \\
        $F \gets F \cup (f,\Build(f(X))$) \Comment*[f]{build on transformed X, store $f$}
    }
    \Return $F$
}

\BlankLine

\Block({$\Build(X)$} \Comment*[f]{recursively build tree (node/leaf)})
{
    \If(\Comment*[f]{termination reached}){$|X| \le p$}{\Return $\Leaf(X)$}
    \Else(\Comment*[f]{split points and recurse})
    {
        $s \gets$ \expr{one of dimensions $D$ at random} \\
        $v \gets $ \expr{\Median of $X$ in dimension $s$} \\
        $(L,R) \gets$ \expr{\Split of $X$ in dimension $s$ at value $v$} \\
        \Return $\Node(c,v,\Build(L),\Build(R))$ \Comment*[f]{build children on $L,R$}\\
    }
}
\caption{\kgeraf: building\label{alg:build}}
\end{algorithm}
\end{document}

which starts producing errors of that nature:

LaTeX Error: Option clash for package geometry.

I have searched for the past 3 hours in the site and applied the numerous answers, but I just can't make this thing compile! Can you?

till to now I never see that someone use geometry packages in beamer slides ... do you can't live without this package? — Zarko, Jun 23 '16 at 23:12
@Zarko I could surely do, if the code manages to compile and the algorithm to be presented in the slides! — gsamaras, Jun 23 '16 at 23:13

score 7 · Accepted Answer · answered Jun 23 '16 at 23:23

You have several errors, I fixed the most important ones. All the \def should better be \newcommand.

Don't load geometry, it doesn't make sense, as beamer already loads it with its own options
Don't load xcolor, rather pass the options to the class as shown
Don't load graphicx
Don't load url, it's already coming with hyperref
Since you need a frame, it must be fragile because you perform macro definitions in it
Use \footnotesize or the algorithm will not fit
Use \protect\kgeraf in captions; probably the definition of \kd should be $k$\mbox{-}d

Here's a working version.

\documentclass[xcolor={dvipsnames,svgnames}]{beamer}

% definitions 
\newcommand{\kd}{\emph{k}-d\space}
\newcommand{\kgeraf}{\kd GeRaF}
\newcommand{\geraf}{{\tt GeRaF}\xspace}
% \newcommand{\att}[1]{\alert{#1}}
\newcommand{\att}[1]{{#1}}

% shortcuts 
\newcommand{\tab}{\hspace*{2em}}
\newcommand{\head}[1]{{\medskip\noindent\emph{#1}}}
\newtheorem{defn}{Definition}
%\usepackage[top=2.5cm,bottom=2.3cm,hmargin=1.5cm]{geometry} % it was hmargin=2.5cm
\usepackage{amssymb,url}
\setcounter{tocdepth}{3}
\usepackage[english]{babel}
\usepackage{multirow}
\usepackage[ruled,vlined,linesnumbered]{algorithm2e}
\usepackage{xspace}
\usepackage{pgfplots}
%\pgfplotsset{compat=1.90} % ize got latex error here
\usepackage{pgfplotstable}
\def\real{{\mathbb R}}

%--------------------------------------------------------------------------------
% Add a period to the end of an abbreviation unless there's one
% already, then \xspace.
\makeatletter
\DeclareRobustCommand\onedot{\futurelet\@let@token\@onedot}
\def\@onedot{\ifx\@let@token.\else.\null\fi\xspace}
%
\def\eg{\emph{e.g}\onedot} \def\Eg{\emph{E.g}\onedot}
\def\ie{\emph{i.e}\onedot} \def\Ie{\emph{I.e}\onedot}
\def\cf{\emph{c.f}\onedot} \def\Cf{\emph{C.f}\onedot}
\def\etc{\emph{etc}\onedot} %\def\vs{\emph{vs}\onedot}
\def\wrt{w.r.t\onedot} \def\dof{d.o.f\onedot}
\def\etal{\emph{et al}\onedot}
\makeatother

%--------------------------------------------------------------------
\begin{document}

\begin{frame}[fragile]
\footnotesize
\begin{algorithm}[H]
\SetFuncSty{textsc}
\SetDataSty{emph}
\newcommand{\expr}[1]{$\langle${#1}$\rangle$}
\newcommand{\commentsty}[1]{{\color{DarkGreen}#1}}
\SetCommentSty{commentsty}
\SetKwComment{Comment}{$\triangleright$ }{}
\SetKwBlock{Block}{function}{}
\SetKwInOut{Input}{input}
\SetKwInOut{Output}{output}
\SetKw{Return}{return}
\SetKwData{Node}{node}
\SetKwData{Leaf}{leaf}
\SetKwFunction{Build}{build}
\SetKwFunction{Variance}{variance}
\SetKwFunction{Median}{median}
\SetKwFunction{Split}{split}

\Input
{pointset $X$, \#trees $m$, \#split-dimensions $t$, max \#points per leaf $p$
}
\Output{randomized \kd forest $F$}
\Begin
{
    $V \gets$ \expr{\Variance of $X$ in every dimension} \\
    $D \gets$ \expr{$t$ dimensions of maximum variance $V$} \\
    $F \gets \emptyset$  \Comment*[f]{forest} \\
    \For{$i \gets 1$ \KwTo $m$}
    {
        $f \gets$ \expr{random transformation}  \Comment*[f]{isometry, shuffling} \\
        $F \gets F \cup (f,\Build(f(X))$) \Comment*[f]{build on transformed X, store $f$}
    }
    \Return $F$
}

\BlankLine

\Block({$\Build(X)$} \Comment*[f]{recursively build tree (node/leaf)})
{
    \If(\Comment*[f]{termination reached}){$|X| \le p$}{\Return $\Leaf(X)$}
    \Else(\Comment*[f]{split points and recurse})
    {
        $s \gets$ \expr{one of dimensions $D$ at random} \\
        $v \gets $ \expr{\Median of $X$ in dimension $s$} \\
        $(L,R) \gets$ \expr{\Split of $X$ in dimension $s$ at value $v$} \\
        \Return $\Node(c,v,\Build(L),\Build(R))$ \Comment*[f]{build children on $L,R$}\\
    }
}
\caption{\protect\kgeraf: building}\label{alg:build}
\end{algorithm}

\end{frame}
\end{document}

You should change your name to latex-God. Thanks! – gsamaras Jun 23 '16 at 23:24 — gsamaras, Jun 23 '16 at 23:24

score 3 · Answer 2 · answered Jun 23 '16 at 23:25

You need to place the algorithm inside a frame. Even then the content is too much to fit on a regular slide. As such, I've placed it inside a box - using an [H] float specifier - and resized it to fit within the frame height.

\documentclass[dvipsnames,svgnames]{beamer}

\usepackage[ruled,vlined,linesnumbered]{algorithm2e}

\begin{document}

\begin{frame}[fragile]
  \resizebox{!}{45mm}{\begin{algorithm}[H]
    \caption{$k$-d GeRaF: building}
    \SetFuncSty{textsc}
    \SetDataSty{emph}
    \newcommand{\expr}[1]{$\langle${#1}$\rangle$}
    \newcommand{\commentsty}[1]{{\color{DarkGreen}#1}}
    \SetCommentSty{commentsty}
    \SetKwComment{Comment}{$\triangleright$ }{}
    \SetKwBlock{Block}{function}{}
    \SetKwInOut{Input}{input}
    \SetKwInOut{Output}{output}
    \SetKw{Return}{return}
    \SetKwData{Node}{node}
    \SetKwData{Leaf}{leaf}
    \SetKwFunction{Build}{build}
    \SetKwFunction{Variance}{variance}
    \SetKwFunction{Median}{median}
    \SetKwFunction{Split}{split}
    \Input
    {pointset $X$, \#trees $m$, \#split-dimensions $t$, max \#points per leaf $p$
    }
    \Output{randomized $k$-d forest $F$}
    \Begin
    {
        $V \gets$ \expr{\Variance of $X$ in every dimension} \\
        $D \gets$ \expr{$t$ dimensions of maximum variance $V$} \\
        $F \gets \emptyset$  \Comment*[f]{forest} \\
        \For{$i \gets 1$ \KwTo $m$}
        {
            $f \gets$ \expr{random transformation}  \Comment*[f]{isometry, shuffling} \\
            $F \gets F \cup (f,\Build(f(X))$) \Comment*[f]{build on transformed X, store $f$}
        }
        \Return $F$
    }
    \BlankLine
    \Block({$\Build(X)$} \Comment*[f]{recursively build tree (node/leaf)})
    {
        \If(\Comment*[f]{termination reached}){$|X| \le p$}{\Return $\Leaf(X)$}
        \Else(\Comment*[f]{split points and recurse})
        {
            $s \gets$ \expr{one of dimensions $D$ at random} \\
            $v \gets $ \expr{\Median of $X$ in dimension $s$} \\
            $(L,R) \gets$ \expr{\Split of $X$ in dimension $s$ at value $v$} \\
            \Return $\Node(c,v,\Build(L),\Build(R))$ \Comment*[f]{build children on $L,R$} \\
        }
    }
  \end{algorithm}}
\end{frame}

\end{document}

I also removed all the superfluous content that doesn't pertain to this example. Additionally, since this is a presentation, some of the general definitions have been inserted in the code to allow it to compile, especially the use of a [fragile] frame.

Note that algorithm2e typically uses \; to denote line-endings.

If you're using a different theme, then you might want to adjust the <height> parameter in \resizebox{!}{<height>}{...}.

OK good Werner. However the other answer came first, but I will upvote you, thank you! — gsamaras, Jun 23 '16 at 23:27

Algorithm with beamer package causing nightmares

2 Answers2

Linked