Download the example data and unpack it into a directory of you choice. The entire dataset has 1.1G and can be obtained from us on request. Download the stimuli and unpack it into a directory of you choice.
To run the code, you need to download the python files and unpack it. The code is written in python 2.7 and requires the following python packages:
python train_classifiers.py -d PATH_TO_DATA -e EOD -c -o OUTPUTDIR -m 1_vs_base -s STIMDIR -b 1000 DATE TYPE
PATH_TO_DATAis the path to the data downloaded before
NO_OF_EOD_CYCLESis the number of eod cycles used for training (must be one of 5, 10, 15, 20, 25, 30, 35, 40, 45, 50)
OUTPUTDIRis the path where the output is stored
DATEis one of the following dates: 2011-06-14, 2011-06-15, 2011-06-20, 2011-06-21, 2011-07-11, 2011-07-13, 2011-07-15, 2011-07-22, 2011-10-17, 2011-10-20, 2011-10-24, 2012-01-11, 2012-01-17, 2012-01-20, 2012-01-26, 2012-01-27, 2012-02-02. The example data is printed in bold.
ampullary. If you want to train classifiers for populations, then
TYPEcan be a colon separated list of types, e.g. p-unit:p-unit:t-unit would train a classifier for populations with 2 p-units and one t-unit.
If you want to produce a figure as in Figure 9 then you need to run the above command for one date and several eod lengths. Then you can run:
python train_classifiers.py -d PATH_TO_DATA -e EOD -c -o NEWOUTPUTDIR -p OLDOUTPUTDIR -m 1_vs_base -s STIMDIR -b 1000 DATE TYPE
NEWOUTPUTDIRis the path where the output is stored (different to the one before, since the script will skip already computed results)
OLDOUTPUTDIRis the path where the output from before was stored. The script uses the results to compute the mutual information curve.
Figure 10 can be reproduced with the script
to work properly, one needs to ensure the following conditions
The datasets provided on this page by A. Stöckl, F. Sinz, J. Benda, J.Grewe are licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License. This license requires that you contact us before you use the data in your own research. In particular, this means that you have to ask for permission if you intend to publish a new analysis performed with this data (no derivative works-clause).
The code for the paper "Encoding of social signals in all three electrosensory pathways of Eigenmannia virescens Data" by A. Stöckl et al. is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.