% \VignetteEngine{knitr::knitr}
% \VignetteIndexEntry{02. Working with R -- slides}

\documentclass[xcolor=dvipsnames]{beamer}

\usepackage{BioconductorSlides}
\hypersetup{colorlinks,linkcolor=,urlcolor=Blue}

\title{Introduction to \R{}}
\author{Martin T.\ Morgan\footnote{\url{mtmorgan@fhcrc.org}}}
\date{27-28 February 2014}

\begin{document}

\maketitle

\section*{R}

\begin{frame}[fragile]{\R}
  \begin{columns}
    \column{.45\textwidth}
    \begin{itemize}
    \item \url{http://r-project.org}
    \item Open-source, statistical programming language; widely used in
      academia, finance, pharma, \ldots
    \item Core language and base packages
    \item Interactive sessions, scripts
    \item $>5000$ contributed packages
    \end{itemize}
    \column{.55\textwidth}
<<R-figures, eval=TRUE, tidy=FALSE, fig.keep="none">>=
## Two 'vectors'
x <- rnorm(1000)
y <- x + rnorm(1000, sd=.5)
## Integrated container
df <- data.frame(X=x, Y=y)
## Visualize
plot(Y ~ X, df)
## Regression; 'object'
fit <- lm(Y ~ X, df)
## Methods on the object
abline(fit) # regression line
anv <- anova(fit) # ANOVA table
@ 
  \end{columns}
\end{frame}

\begin{frame}{Programming \R}
  \begin{enumerate}
  \item Packages: loading, installing
  \item Help
  \item Scripts \& reproducible research
  \item Functions
  \item Debugging and measuring performance
  \end{enumerate}
\end{frame}

\begin{frame}[fragile]{1. Packages}
  Already installed packages
<<installed-packages>>=
library(parallel)
@
%% 
New packages from \emph{repositories} such as
\href{http://cran.r-project.org}{CRAN} and
\href{http://bioconductor.org/}{Bioconductor}
\begin{itemize}
\item \Rcode{biocLite()} to install, including \emph{dependencies}
\item Occasional problems when a package depends on third-party
  software installation
\end{itemize}
<<biocLite, eval=FALSE>>=
source("http://bioconductor.org/biocLite.R")
biocLite("IRanges")
library("IRanges")
@ 
%% 
Other repositories: \texttt{R-forge}, \texttt{github}, \ldots
\end{frame}

\begin{frame}[fragile]{Packages (cont.)}
  What packages are loaded?
<<search>>=
head(search(), 3)
@ 
  What functions are provided by a package?
<<package-help, eval=FALSE>>=
help(package="IRanges")
@   
  How does \R{} find symbols, e.g,. \Rcode{sin}?
  \begin{itemize}
  \item Look in \Rcode{.GlobalEnv}, then proceed down search path
  \item Specify package with \Rcode{base::sin}
  \end{itemize}
\end{frame}

\begin{frame}[fragile]{2. Help}
\begin{knitrout}
\definecolor{shadecolor}{rgb}{0.969, 0.969, 0.969}\color{fgcolor}\begin{kframe}
\begin{alltt}
\hlkwd{help.start}\hlstd{()}
\hlkwd{? }\hlstd{data.frame}
\hlkwd{? }\hlstd{anova}
\hlkwd{? }\hlstd{anova.lm}  \hlcom{# anova generic, method for class lm}
\hlkwd{class ? }\hlstd{DNAStringSet}
\hlkwd{method ? }\hlstr{"alphabetFrequency,DNAStringSet"}
\hlkwd{vignette}\hlstd{(}\hlstr{"GenomicRangesIntroduction"}\hlstd{,} \hlstr{"GenomicRanges"}\hlstd{)}
\hlkwd{help}\hlstd{(}\hlkwc{package} \hlstd{=} \hlstr{"Biostrings"}\hlstd{)}
\hlkwd{RShowDoc}\hlstd{(}\hlstr{"R-intro"}\hlstd{)}
\end{alltt}
\end{kframe}
\end{knitrout}
\end{frame}

\begin{frame}{3. Scripts, functions, and reproducible research}
  \begin{enumerate}
  \item Write  simple scripts of \R{} code \texttt{my\_analysis.R}
  \item Implement common operations as functions.
  \item `Markdown' with \R{} code embedded in surrounding text
    \texttt{my\_analysis.Rmd}
  \item Packages! -- \Rcode{package.skeleton()}
  \item Version control!
  \end{enumerate}
\end{frame}

\begin{frame}{4. Favorite functions}
  \begin{columns}[t]

    \column{.5\textwidth}

    \begin{description}\itemsep0pt\parskip0pt
    \item[\Rfunction{dir}, \Rfunction{read.table}, \Rfunction{scan}]
      List files; input data.
    \item[\Rfunction{c}, \Rfunction{factor}, \Rfunction{data.frame},
      \Rfunction{matrix}] Create vectors, etc.
    \item[\Rfunction{summary}, \Rfunction{table}, \Rfunction{xtabs}]
      Summarize or cross-tabulate data.
    \item[\Rfunction{t.test}, \Rfunction{lm}, \Rfunction{anova}]
      Compare two or several groups.
    \item[\Rfunction{dist}, \Rfunction{hclust}, \Rfunction{heatmap}]
      Cluster data.
    \item[\Rfunction{plot}] Plot data.
    \item[\Rfunction{ls}, \Rfunction{library}] List objects; attach
      packages.
    \end{description}

    \column{.5\textwidth}

    \begin{description} \itemsep0pt\parskip0pt
    \item[\Rfunction{lapply}, \Rfunction{sapply}, \Rfunction{mapply}]
      Apply function to elements of lists.
    \item[\Rfunction{match}, \Rfunction{\%in\%}] find elements of one
      vector in another.
    \item[\Rfunction{split}, \Rfunction{cut}] Split or cut vectors.
    \item[\Rfunction{strsplit}, \Rfunction{grep}, \Rfunction{sub}] Operate
      on character vectors.
    \item[\Rfunction{biocLite}] Install a package from an on-line
      repository.
    \end{description}

  \end{columns}
\end{frame}

\begin{frame}{5. Debugging and measuring performance}
  Debugging
  \begin{itemize}
    \item \Rfunction{traceback()}: what went wrong?
    \item \Rfunction{debug()}: step through a function.
    \item \Rfunction{browser()}: insert a break-point in your own
      function / script.  Help debug errors.
  \end{itemize}
\end{frame}

\begin{frame}{Debugging and measuring performance (cont.)}
  Performance
  \begin{itemize}
  \item \Rfunction{all.equal()}, \Rfunction{identical()} to compare
    values.
  \item \Rfunction{system.time()} to measure how long evaluation takes.
  \item \Rpkg{microbenchmark} to compare times for different
    functions
  \item \Rfunction{Rprof()} to summarize time in each function call,
    \href{https://github.com/hadley/lineprof}{lineprof} to profile
    each line of code
  \end{itemize}
\end{frame}

\end{document}