Changeset 348

Show
Ignore:
Timestamp:
06/06/08 16:13:45 (6 months ago)
Author:
apdavison
Message:

Freiburg lecture: Wrote most of the section on PyNN backends

Files:

Legend:

Unmodified
Added
Removed
Modified
Copied
Moved

  • doc/Freiburg2008/backends.tex

    r321 r348  
    88\frametitle{NEST} 
    99 
    10 \end{frame} 
    11  
    12  
    13 \begin{frame} % -------------------------------------------------------------------- 
     10%NEST = "Neural Simulation Tool" 
     11 
     12\url{http://www.nest-initiative.org/} 
     13 
     14\begin{block}{} 
     15\begin{itemize} 
     16\item Best suited for large networks of spiking point-neuron models. 
     17\item Focus on efficiency and on flexible design of network architectures 
     18\item Adding new neuron models is not easy (requires writing C++ and recompiling) 
     19\item Runs on single processors, clusters, BlueGene 
     20\item Choice of two scripting languages: SLI and Python 
     21\end{itemize} 
     22\end{block} 
     23 
     24\begin{block}{\footnotesize Developers:} 
     25\scriptsize 
     26Marc-Oliver Gewaltig (Honda Research Institute Europe), 
     27Markus Diesmann (RIKEN Brain Science Institute), 
     28Abigail Morrison (RIKEN Brain Science Institute), 
     29Hans Ekkehard Plesser (Norwegian University of Life Sciences), 
     30Jochen Eppler (Honda Research Institute Europe), 
     31Moritz Helias (BCCN Freiburg), 
     32$\ldots$ 
     33\end{block} 
     34 
     35\end{frame} 
     36 
     37\begin{frame} % -------------------------------------------------------------------- 
     38\frametitle{NEST} 
     39\framesubtitle{Download} 
     40 
     41\begin{block}{NEST 1.0} 
     42\small 
     43\url{http://www.nest-initiative.org/index.php/Software:Download} 
     44 
     45\url{http://www.nest-initiative.org/index.php/Image:Pynest1.tgz} 
     46\end{block} 
     47 
     48\begin{block}{NEST 2.0} 
     49\small 
     50available on request from Marc-Oliver Gewaltig (marc-oliver.gewaltig@honda-ri.de) 
     51 
     52(PyNEST included). 
     53\end{block} 
     54 
     55\end{frame} 
     56 
     57 
     58\begin{frame}[fragile] % -------------------------------------------------------------------- 
    1459\frametitle{NEST} 
    1560\framesubtitle{Installation} 
    1661 
     62 
     63 
     64\begin{verbatim} 
     65./configure --enable-pynest --with-mpi 
     66make 
     67make install 
     68\end{verbatim} 
     69 
     70\textit{(NEST 2.0)} 
     71 
     72 
    1773\end{frame} 
    1874 
    …… 
    2177\framesubtitle{Documentation and resources} 
    2278 
    23 \end{frame} 
    24  
    25 \begin{frame} % -------------------------------------------------------------------- 
    26 \frametitle{NEST} 
    27 \framesubtitle{PyNEST} 
     79\begin{block}{Documentation} 
     80\footnotesize 
     81\url{http://www.nest-initiative.org/index.php/Software:Documentation} 
     82 
     83\vspace{2ex} 
     84 
     85\url{http://www.scholarpedia.org/article/NEST_(NEural_Simulation_Tool)} 
     86\end{block} 
     87 
     88\begin{block}{Mailing list} 
     89\footnotesize 
     90\url{http://www.nest-initiative.org/index.php/Software:Mailinglists} 
     91\end{block} 
     92 
     93\end{frame} 
     94 
     95\begin{frame}[fragile] % -------------------------------------------------------------------- 
     96\frametitle{NEST} 
     97\framesubtitle{Example} 
     98 
     99\begin{verbatim} 
     100import nest 
     101 
     102neuron = nest.Create("iaf_neuron") 
     103 
     104input = nest.Create("dc_generator") 
     105nest.SetStatus(input, {"amplitude": 550.0}) 
     106 
     107vm = nest.Create("voltmeter") 
     108 
     109nest.Connect(input, neuron) 
     110nest.Connect(vm, neuron) 
     111 
     112nest.Simulate(100.0) 
     113\end{verbatim} 
    28114 
    29115\end{frame} 
    …… 
    35121\frametitle{NEURON} 
    36122 
    37 \end{frame} 
    38  
    39  
    40 \begin{frame} % -------------------------------------------------------------------- 
     123\url{http://www.neuron.yale.edu/} 
     124 
     125\begin{block}{} 
     126\begin{itemize} 
     127\item Best suited for networks of compartmental models 
     128\item Highly flexible and extensible 
     129\item Widely used (739 papers using NEURON as of 07/03/08) 
     130\item Adding new neuron models uses interpreter or GUI  
     131\item Runs on single processors, clusters, BlueGene 
     132\item Choice of two scripting languages: hoc and Python 
     133\end{itemize} 
     134\end{block} 
     135 
     136\begin{block}{\footnotesize Developers:} 
     137\scriptsize 
     138Michael Hines (Yale University), $\ldots$ 
     139\end{block} 
     140 
     141 
     142\end{frame} 
     143 
     144\begin{frame} % -------------------------------------------------------------------- 
     145\frametitle{NEURON} 
     146\framesubtitle{Download} 
     147 
     148\small 
     149\url{http://www.neuron.yale.edu/neuron/install/install.html} 
     150 
     151\end{frame} 
     152 
     153\begin{frame}[fragile] % -------------------------------------------------------------------- 
    41154\frametitle{NEURON} 
    42155\framesubtitle{Installation} 
    43156 
     157Pre-built versions for Windows, Max OS X, Linux (RPM) 
     158 
     159\vspace{2ex} 
     160 
     161or compile from source. 
     162 
     163\begin{verbatim} 
     164configure --with-nrnpython --with-mpi 
     165make 
     166make install 
     167\end{verbatim} 
     168 
     169 
    44170\end{frame} 
    45171 
    …… 
    48174\framesubtitle{Documentation and resources} 
    49175 
    50 \end{frame} 
    51  
    52 \begin{frame} % -------------------------------------------------------------------- 
    53 \frametitle{NEURON} 
    54 \framesubtitle{nrnpython} 
    55  
     176\begin{block}{Documentation} 
     177\small 
     178http://www.neuron.yale.edu/neuron/docs/docs.html 
     179\end{block} 
     180 
     181\begin{block}{Users' Group Forum} 
     182\small 
     183http://www.neuron.yale.edu/phpBB2/ 
     184\end{block} 
     185 
     186\begin{block}{ModelDB} 
     187\small 
     188http://senselab.med.yale.edu/modeldb/ 
     189\end{block} 
     190 
     191\end{frame} 
     192 
     193\begin{frame}[fragile] % -------------------------------------------------------------------- 
     194\frametitle{NEURON} 
     195\framesubtitle{Example} 
     196 
     197\begin{verbatim} 
     198import neuron, nrn 
     199 
     200soma = nrn.Section() 
     201soma.insert("hh") 
     202 
     203input = neuron.IClamp(soma, 0.5) 
     204input.dur = 100 
     205input.amp = 0.1 
     206 
     207neuron.init() 
     208neuron.run(100.0) 
     209\end{verbatim} 
    56210\end{frame} 
    57211 
    …… 
    63217\frametitle{PCSIM} 
    64218 
    65 \end{frame} 
    66  
    67  
    68 \begin{frame} % -------------------------------------------------------------------- 
     219\url{http://www.lsm.tugraz.at/pcsim/} 
     220 
     221\begin{block}{} 
     222\begin{itemize} 
     223\item Best suited for large networks of spiking point-neuron models. 
     224\item Focus on efficiency 
     225\item Adding new neuron models is not easy (requires writing C++ and recompiling) 
     226\item Runs on single processors, clusters 
     227\item Python 
     228\end{itemize} 
     229\end{block} 
     230 
     231\begin{block}{\footnotesize Developers:} 
     232\scriptsize 
     233Dejan Pecevski, Thomas Natschl\"ager (Graz University of Technology) 
     234\end{block} 
     235 
     236\end{frame} 
     237 
     238 
     239\begin{frame}[fragile] % -------------------------------------------------------------------- 
    69240\frametitle{PCSIM} 
    70 \framesubtitle{Installation} 
     241\framesubtitle{Download and Installation} 
     242 
     243\url{http://sourceforge.net/projects/pcsim/} 
     244\vspace{1ex} 
     245 
     246A long list of dependencies:\\ 
     247{\footnotesize CMake, Doxygen, elementtree, Boost C++, MPI, cppunit, GSL, gccxml, pygccxml, Py++} 
     248 
     249\vspace{1ex} 
     250See \texttt{\small HowTo-Install-PCSIM-Dependencies-Linux.txt} or \texttt{\small HowTo-Install-PCSIM-Dependencies-Windows.txt} in the source distribution. 
     251 
     252\vspace{2ex} 
     253After that, it's easy: 
     254\begin{verbatim} 
     255python setup.py install 
     256python setup.py test 
     257\end{verbatim} 
    71258 
    72259\end{frame} 
    …… 
    76263\framesubtitle{Documentation and resources} 
    77264 
    78 \end{frame} 
    79  
    80 \begin{frame} % -------------------------------------------------------------------- 
     265\begin{block}{User Manual} 
     266\footnotesize 
     267\url{http://www.lsm.tugraz.at/pcsim/usermanual/html/index.html} 
     268\end{block} 
     269 
     270\begin{block}{Python class reference} 
     271\footnotesize 
     272\url{http://www.lsm.tugraz.at/pcsim/pyclassreference/html/index.html} 
     273\end{block} 
     274 
     275\begin{block}{User forums/Mailing list} 
     276\footnotesize 
     277http://sourceforge.net/projects/pcsim/ 
     278\end{block} 
     279 
     280\end{frame} 
     281 
     282\begin{frame}[fragile] % -------------------------------------------------------------------- 
    81283\frametitle{PCSIM} 
    82 \framesubtitle{PyPCSIM} 
     284\framesubtitle{Example} 
     285 
     286\begin{verbatim} 
     287import pypcsim 
     288 
     289net = pypcsim.SingleThreadNetwork() 
     290 
     291neuron = net.create(pypcsim.LifNeuron(Iinject=1e-8)) 
     292 
     293vm = net.record(neuron, "Vm", pypcsim.AnalogRecorder()) 
     294 
     295net.simulate(0.1) 
     296\end{verbatim} 
    83297 
    84298\end{frame} 
    …… 
    92306\frametitle{Brian} 
    93307 
    94 \end{frame} 
    95  
    96  
    97 \begin{frame} % -------------------------------------------------------------------- 
     308\url{http://brian.di.ens.fr/} 
     309 
     310\begin{block}{} 
     311\begin{itemize} 
     312\item Best suited for networks of spiking point-neuron models. 
     313\item Focus on flexibility and extensibility 
     314\item Adding new neuron models is very easy (just type the equations) 
     315\item Runs on single processors 
     316\item Pure Python 
     317\end{itemize} 
     318\end{block} 
     319 
     320\begin{block}{\footnotesize Developers:} 
     321\scriptsize 
     322Romain Brette, Dan Goodman (ENS, Paris). 
     323\end{block} 
     324 
     325\end{frame} 
     326 
     327 
     328\begin{frame}[fragile] % -------------------------------------------------------------------- 
    98329\frametitle{Brian} 
    99 \framesubtitle{Installation} 
     330\framesubtitle{Download and Installation} 
     331 
     332\url{https://gforge.inria.fr/frs/?group_id=1069} 
     333 
     334\vspace{2ex} 
     335 
     336Dependencies (you should mostly have anyway):\\ 
     337{\small numpy, scipy, sympy, pylab} 
     338 
     339\vspace{2ex} 
     340 
     341Pre-built installer for Windows 
     342 
     343\vspace{2ex} 
     344 
     345or install from source: 
     346 
     347\begin{verbatim} 
     348python setup.py install 
     349\end{verbatim} 
    100350 
    101351\end{frame} 
    …… 
    105355\framesubtitle{Documentation and resources} 
    106356 
     357\begin{block}{Documentation} 
     358\url{http://brian.di.ens.fr/docs/} 
     359\end{block} 
     360 
     361\begin{block}{Support} 
     362\url{http://groups.google.fr/group/briansupport} 
     363\end{block} 
     364 
    107365\end{frame} 
    108366 
    109367\begin{frame} % -------------------------------------------------------------------- 
    110368\frametitle{Brian} 
    111 \framesubtitle{Examples} 
    112  
    113 \end{frame} 
    114  
     369\framesubtitle{Example} 
     370 
     371 
     372 
     373\end{frame} 
     374