| hypernad.pdf | (6705 kb) | HyperNAD result diagrams |
| hypernad.zip | (62 kb) | scripts and other source files |
| hypernad.tar.gz | (54 kb) |
HyperNAD (Hypergraph-based Neuron Assembly Detection) is a program to analyze parallel neuronal spike trains for assembly activity. It allows for selective participation in the individual coincident firing events by using a hypergraph-based interpretation of the (filtered) pattern set returned by CoCoNAD, which is then processed by constructing a reduction sequence to find a densest sub-hypergraph.
The document hypernad.pdf contains the result diagrams
for the complete set of experiments with HyperNAD to find synchronous
spiking events with selective participation in parallel neural spike
trains, that were conducted for the paper
[Borgelt et al. 2015].
Only few of these diagrams are contained in the paper due to a
lack of space.
The archives hypernad.{zip,tar.gz} contain scripts and
other source files, with which the experiments were conducted and
the document with the result diagrams was created. Call the script
hypernad.py without any arguments to obtain a help message
that shows the invocation and the available options.
Details of the method can be found in the paper [Borgelt et al. 2015]. Details about the CoCoNAD pattern mining approach can be found in the papers [Borgelt and Picado-Muiño 2013] and [Picado-Muiño and Borgelt 2014]. Details about the pattern spectrum filtering method that is used to preprocess the found patterns (although for a setting with time-binning) can be found in the papers [Picado-Muiño et al. 2013] and [Torre et al. 2013]. An alternative approach to deal with selective participation, which, however, uses time-binning (which loses information), can be found in [Borgelt et al. 2011].
Note that the scripts etc. were developed on/for a GNU/Linux
system (Ubuntu 14.04 or later) and thus are directly executable on such
a system or a similar one (that is, some other GNU/Linux distribution).
Although at least most of the Python scripts should also be working
on a Windows system (with the possible exception of the parallelization
scripts), most of the other scripts (like the run script,
which is the main control script, and the makefile, which
controls generating the diagrams from the result data) may need
porting to batch files or something similar.
On a GNU/Linux system, the following software needs to be installed to run the experiments:
cocoexp package will also work
without this extension module, namely by falling back on a pure
Python replacement, which, however, is slower by a factor of
about 40 or more),pdflatex program,mptopdf command,bash,
awk, tar etc.) and are easy to install
otherwise.On such a system the experiments can be run by simply calling
the main script run (in the directory hypernad)
on the command line, which does everything. The execution of the
experiments exploits 4-fold parallelization, thus making full use of the
quadcore processors basically all modern computers are equipped with.
The progress of the experiments can be followed on the command line,
to which regular progress messages are written. Once all experiments
are completed (which, even on a modern computer system, can take several
days, mainly because of the huge number of individual experimental runs,
namely in the hundreds of thousands), the result diagrams are created
and compiled into the final documents, which are also directly
available above.