\documentclass[english]{article}
\usepackage[T1]{fontenc}
\usepackage[latin9]{inputenc}
\usepackage{calc}
\usepackage{babel}
\usepackage{hyperref}
\usepackage{url}
\begin{document}

% \VignetteIndexEntry{MuToss Quick Start Guide}

\title{$\mu$TOSS Quick Start Guide}


 \author{\scriptsize {Gilles Blanchard}, % \url{gilles.blanchard@wias-berlin.de}
 \scriptsize Weierstrass Institute for Applied Analysis and Stochastics Berlin
 \and
 \scriptsize {Thorsten Dickhaus}, % \url{dickhaus@math.hu-berlin.de}
 \scriptsize Humboldt-University Berlin
 \and
 \scriptsize {Niklas Hack}, % \url{niklas.hack@meduniwien.ac.at}
 \scriptsize Medical University of Vienna
 \and
 \scriptsize {Frank Konietschke}, % \url{Frank.Konietschke@medizin.uni-goettingen.de}
 \scriptsize Georg-August-University G\"ottingen
 \and
 \scriptsize {Kornelius Rohmeyer}, % \url{rohmeyer@biostat.uni-hannover.de}
 \scriptsize Leibniz University Hannover
 \and
 \scriptsize {Jonathan Rosenblatt}, % \url{john.ros@gmail.com}
 \scriptsize Tel Aviv University
 \and
 \scriptsize {Marsel Scheer}, % \url{mscheer@ddz.uni-duesseldorf.de}
 \scriptsize German Diabetes Center
 \and
 \scriptsize {Wiebke Werft}, % \url{w.werft@Dkfz-Heidelberg.de}
 \scriptsize German Cancer Research Center
 }

\maketitle

\begin{abstract}
$\mu$TOSS is an \verb=R= package providing an open source, easy-to-extend platform for
multiple hypothesis testing (MHT), one of the most active research fields in statistics over the last 10-15 years. Its first motivation is to establish a
common platform and standardization for MHT procedures at large. The
$\mu$TOSS software has been designed and written in the framework of a ``Harvest Programme'' call of the PASCAL2 European research network. Basically,
it consists of the two R packages \verb=mutoss= and \verb=mutossGUI=. For researchers, it features a convenient unification of interfaces for MHT procedures (including standardized functions to access existing specific MHT \verb=R= packages such as \verb=multtest= and \verb=multcomp=, as well as recent MHT procedures that are not available elsewhere)  and
helper functions facilitating the setup of benchmark simulations for comparison of competing methods. For end users, a graphical user interface and an online user's guide for finding appropriate methods for a given specification of the multiple testing problem is included. Ongoing maintenance and subsequent extensions will aim at establishing $\mu$TOSS as a state of the art in statistical computing for MHT.
\end{abstract}

\section{Introduction}	

The $\mu$TOSS packages allow the user to discover, apply and compare
multiple testing procedure and multiple interval estimation procedures.

The $\mu$TOSS packages include a corpus of functions implementing
and integrating these procedures and a GUI. These are found in the
mutoss and mutossGUI packages respectively.


\section{$\mu$TOSS Rationale}

The rationale behind the $\mu$TOSS packages is two-fold.

It is aimed at allowing statisticians to discover, apply and compare
standard and custom multiplicity controlling procedures. This is achieved
by the mutoss package.

It is also aimed at the researcher wishing to Analise new data or
reproduce published results. This is accomplished by the mutossGUI
package.

%
\framebox{\begin{minipage}[t]{1\columnwidth}%
At the time of release, the package has only undergone basic testing.
This being the case, we recommend new data to be analyzed with standard
software alongside $\mu$TOSS. This is planned to change in future
releases.%
\end{minipage}}


\section{System Requirements}


\subsection{mutoss Package}

The package will run on any machine running R with recommended version
2.8 and above.


\subsection{mutossGUI package}

On top of the mutoss package requirements, Java Run time Environment
ver 5 and above is needed.


\section{GUI Work flow}

Download and install the \textit{mutossGUI} package. The GUI should
start automatically. Others wise load it with

\texttt{>mutossGUI()}


\subsection{Testing of Hypotheses}

If you have already a vector of p-values start at step (5). 
\begin{enumerate}
\item Load the raw data (assumed to be a \textit{data.frame} object) using
the \textbf{Data} button.
\item Specify the model type and explanatory variables using the \textbf{Model}
button.\\
For linear contrasts choose \textbf{Single endpoint in k groups}.\\
For applying the same model to many response variables choose \textbf{Multiple
(linear) regression.}
\item Define model by choosing response and expalnatory variables.
\item Define the hypotheses of interest by specifying the contrasts using
the \textbf{Hypotheses} button.
\item Insert p-values using the \textbf{p-Value} button or calculate them
following the previous steps.
\item Choose the error type to control using the \textbf{Error Rate} button.
\item Use the \textbf{Adjusted p-Values} to calculate the procedure specific
adjusted p-values (you will be propted for additional options when
necesary) or choose \textbf{Rejeted} to apply the procedure and reject
hypotheses.
\item Visualize results by choosing the \textbf{Info} option in the \textbf{Adjusted
p-Values} or \textbf{Rejected} buttons.
\item Save the output as an R object using the \textbf{File->Export MuToss
Object to R} option.
\end{enumerate}
Further analysis is now possible using the compareMutoss functions
or other R functionality.


\subsection{Interval Estimations}

Steps 1-4 are identical to the hypothesis testing work flow.
\begin{enumerate}
\item Load the raw data (assumed to be a \textit{data.frame} object) using
the \textbf{Data} button.
\item Specify the model type and explanatory variables using the \textbf{Model}
button.\\
For linear contrasts choose \textbf{Single endpoint in k groups}.\\
For applying the same model to many response variables choose \textbf{Multiple
(linear) regression.}
\item Define model by choosing response and expalnatory variables.
\item Define the contrasts of interest by specifying the contrasts using
the \textbf{Hypotheses} button.
\item Choose the error type to control using the \textbf{Error Rate} button.
\item Use the \textbf{Confidence Intervals }to compute confidence intervals
on parameters of interest. 
\item Visualize results by choosing the \textbf{Info} option in the \textbf{Confidence
Intervals} button.
\item Save the output as an R object using the \textbf{File->Export MuToss
Object to R} option.
\end{enumerate}
Further analysis is now possible using R functionality.


\section{Command Line Work Flow}

Download and install the \textit{mutoss} package to access the different
procedures in the package (note mutossGUI is not needed for this purpose).
A list can presented using\\
\texttt{>help(package='mutoss')}

To work with these elementary functions, just use them as any other
R function. Seee inline help for further details.

To use these functions to read and write into Mutoss S4 class objects
use the \textit{mutoss.apply( )} function. See the inline help of
the function for further details.


\section{Glossary}
\begin{description}
\item [{Hypotheses-Testing-Procedures}] The corpus of procedures for testing
multiple statistical hypotheses.
\item [{Interval-Estimating-Procedures}] The corpus of procedures for constructing
interval estimates for multiple parameters.
\item [{p-Value-Procedures}] The corpus of (multiple) hypotheses testing
procedures which rely on the marginal p-values of each hypothesis
(and do not require the original data and model). This procedures
might possibly require additional information such as logical relations
between procedures, a qualitative description of the probabilistic
relation between test statistics etc.
\item [{Data-Procedures}] The corpus of (multiple) testing procedures which
require the original response variables, the explanatory variables
(model) and the parameters of interest (contrasts). \\
These procedures can be seen as p-value-procedures with a specific
relation between test-statistics which is derived from the model and
the contrasts.
\item [{Error-Type}] The type of error a procedure aims to control. This
can be a hypothesis testing error rate (FWER. FDR,...) or an interval
estimation error rate (simultanous coverage, false coverage rate,...).
\item [{Error-Rate}] The allowed rate of the \textit{Error Type}. 
\end{description}

\end{document}