Skip to content

Commit

Permalink
Changed all 'code' blocks to 'lstlisting' blocks
Browse files Browse the repository at this point in the history
  • Loading branch information
@drewconway
drewconway committed Jun 16, 2010
1 parent fff8a32 commit 8085b88
Show file tree
Hide file tree
Showing 4 changed files with 216 additions and 228 deletions.
Binary file modified 2-Why_Do_SNA_with_NX/module_II.pdf
View file
Binary file not shown.
76 changes: 38 additions & 38 deletions 2-Why_Do_SNA_with_NX/module_II.tex
View file
Expand Up @@ -8,7 +8,7 @@

\input{../tex/workshop_style.tex}

\title{Module II - Why do SNA in \texttt{NetworkX}}
\title{Module II - Why do SNA in NetworkX}
\author{Drew Conway and Aric Hagberg}
% \institute{\includegraphics[width = 4cm]{../images/logos/500px-NYU_logo.png}}
\date{June 29, 2010}
Expand All @@ -24,16 +24,16 @@
Speed, Scalability \& Graph Types
\begin{itemize}
\item Why speed and scalability matter
\item Comparing \texttt{NetworkX} to other SNA tools
\item What can be a ``graph'' in \texttt{NetworkX}
\item Comparing NetworkX to other SNA tools
\item What can be a ``graph'' in NetworkX
\end{itemize}
\uncover<2->{How \texttt{NetworkX} complements Python's scientific computing suite
\uncover<2->{How NetworkX complements Python's scientific computing suite
\begin{itemize}
\item \texttt{SciPy}/\texttt{NumPy}
\item \texttt{Matplotlib}
\item SciPy/NumPy
\item Matplotlib
\item GraphViz
\end{itemize}}
\uncover<3->{Getting data in and out of \texttt{NetworkX}
\uncover<3->{Getting data in and out of NetworkX
\begin{itemize}
\item I/O basics
\item Pulling non-local data
Expand Down Expand Up @@ -77,7 +77,7 @@ \subsection{Network data size and tool scalabilty} % (fold)
\end{tabular}
\uncover<3->{\begin{columns}
\column{.5\textwidth}
\texttt{NetworkX} is designed to handle data sets of the scale being generated today
NetworkX is designed to handle data sets of the scale being generated today
\begin{itemize}
\scriptsize{\item 10M's nodes and 100M's edges
\item Read network data from local files, or from external sources
Expand All @@ -104,35 +104,35 @@ \subsection{Supporting non-traditional graph types} % (fold)
\item Networks are collections of nodes and edges
\item Nodes are static integers or strings, and edges are binary or continuous values
\end{itemize}
\texttt{NetworkX} can represent \textbf{ANY} relationship supported by Python data types
NetworkX can represent \textbf{ANY} relationship supported by Python data types
\vspace{2mm}
\begin{columns}
\column{.3\textwidth}
\small{Suppose we had data, or a data generating process, that was a time-series
\begin{itemize}
\item Current tools need kludges or hacks to add this data
\item In \texttt{NetworkX}, we simply use the built-in Python \texttt{datetime} package to create a network of time-stamps
\item In NetworkX, we simply use the built-in Python \texttt{datetime} package to create a network of time-stamps
\end{itemize}}
\column{.7\textwidth}
\begin{block}{}
\begin{code}
\tiny{G=nx.DiGraph()
\alert<2>{# Create datetime object nodes
\tiny{\begin{lstlisting}[language=Python]
G=nx.DiGraph()
# Create datetime object nodes
for v in xrange(num_nodes):
G.add_node(datetime.now())
time_nodes=G.nodes()}
\alert<3>{# Add edges with `time' attribut
time_nodes=G.nodes()
# Add edges with `time' attribut
for i in xrange(num_nodes):
draws=random.uniform(0,1,num_nodes)
for j in xrange(num_nodes):
if i!=j and draws[j]<=p:
G.add_edge(time_nodes[i],time_nodes[j],time=datetime.now())
...}
\alert<4>{# target source datetime_created
...
# target source datetime_created
2010-05-25 13:38:42.515323 2010-05-25 13:38:42.515492
\{`time': datetime.datetime(2010, 5, 25, 13, 38, 42, 515752)\}
...}}
\end{code}
...
\end{lstlisting}}
\end{block}
\end{columns}
\end{frame}
Expand Down Expand Up @@ -198,21 +198,21 @@ \subsection{Working with GraphViz} % (fold)
\label{sub:working_with_graphviz}

\begin{frame}[fragile]
\frametitle{Exporting to GraphViz in \texttt{NetworkX}}
\texttt{NetworkX} is designed to be an open-source all-purpose network manipulation and analysis tool
\frametitle{Exporting to GraphViz in NetworkX}
NetworkX is designed to be an open-source all-purpose network manipulation and analysis tool
\begin{itemize}
\item Historically, the focus has not been on visualization
\end{itemize}
While there are several options for visualization in \texttt{NetworkX}, perhaps the best is its ability to read and write \texttt{GraphViz} files
While there are several options for visualization in NetworkX, perhaps the best is its ability to read and write \texttt{GraphViz} files
\begin{itemize}
\item GraphViz is an open-source tool designed specifically for drawing graphs from the DOT language
\item \texttt{NetworkX} works directly with GV using the \texttt{pygraphviz} package
\item NetworkX works directly with GV using the \texttt{pygraphviz} package
\end{itemize}
\begin{columns}
\column{.68\textwidth}
\begin{block}{}
\begin{code}
\tiny{\alert<2>{# Load Sampson monastery data from edgelist
\tiny{\begin{lstlisting}[language=Python]
# Load Sampson monastery data from edgelist
>>> g2=nx.read_edgelist("samp_like_el.txt",create_using=nx.DiGraph())
>>> nx.info(g2)
Name:
Expand All @@ -221,16 +221,16 @@ \subsection{Working with GraphViz} % (fold)
Number of edges: 55
Average in degree: 3.0556
Average out degree: 3.0556}
\alert<3>{# Convert to pygraphviz type
>>> g2_gv=nx.to_agraph(g2)}
\alert<4>{# Output DOT file and draw using dot layout
# Convert to pygraphviz type
>>> g2_gv=nx.to_agraph(g2)
# Output DOT file and draw using dot layout
>>> g2_gv.write(``1samp_like_dot.dot'')
>>> g2_gv.draw(``samp_like.png'',prog=``dot'')}}
\end{code}
>>> g2_gv.draw(``samp_like.png'',prog=``dot'')
\end{lstlisting}}
\end{block}
\column{.32\textwidth}
\begin{center}
\uncover<5->{\includegraphics[height=4cm]{../images/networks/samp_like.png}}
\uncover<2->{\includegraphics[height=4cm]{../images/networks/samp_like.png}}
\end{center}
\end{columns}
\end{frame}
Expand All @@ -246,14 +246,14 @@ \subsection{Data I/O Basics} % (fold)
\label{sub:data_i_o_basics}

\begin{frame}[fragile]
\frametitle{Getting local data into \texttt{NetworkX}}
Getting data into \texttt{NetworkX} is as simple as a single line of code:
\frametitle{Getting local data into NetworkX}
Getting data into NetworkX is as simple as a single line of code:
\begin{block}{Loading local data file}
\begin{code}
\begin{lstlisting}[language=Python]
>>> G=read_edgelist("my_data.txt")
\end{code}
\end{lstlisting}
\end{block}
\uncover<2->{Like many other network analysis platforms, \texttt{NetworkX} can parse a wide variety of network data types
\uncover<2->{Like many other network analysis platforms, NetworkX can parse a wide variety of network data types
\begin{center}
\begin{tabular}{cll}
\multicolumn{3}{c}{\Large{Readable and Writeable Formats in NX}} \\ \hline \hline
Expand Down Expand Up @@ -296,7 +296,7 @@ \subsection{Loading data from external sources} % (fold)
\begin{itemize}
\item Each service provides different relational context
\item Data formats are generally JSON, Atom, XML, or some combination
\item Python has built-in parsers for all of these data types, which can easily be represented in \texttt{NetworkX}}
\item Python has built-in parsers for all of these data types, which can easily be represented in NetworkX}
\end{itemize}}
\uncover<3->{\alert{Next, we will go over an example of building network data using Google's SocialGraph API}}
\end{frame}
Expand Down Expand Up @@ -330,7 +330,7 @@ \subsection{Loading data from external sources} % (fold)

\begin{frame}[plain]
\frametitle{Module II Review}
Why use \texttt{NetworkX} to do SNA?
Why use NetworkX to do SNA?
\begin{enumerate}
\uncover<2->{\item Unlike many other tools, NX is designed to handle data on a scale relevant to modern problems}
\uncover<3->{\item Most of the core algorithms in NX rely on extremely fast legacy code}
Expand Down
Binary file modified 4-IV_Basic_Analysis/module_IV.pdf
View file
Binary file not shown.

0 comments on commit 8085b88

Please sign in to comment.