Source code of simulations and analyses from Maith, O., Baladron, J., Einhäuser, W., & Hamker, F. H. (2023). Exploration behavior after reversals is predicted by STN-GPe synaptic plasticity in a basal ganglia model. Submitted to iScience.
- Oliver Maith (oliver.maith@informatik.tu-chemnitz.de)
- Javier Baladron (javier.baladron-pezoa@informatik.tu-chemnitz.de)
- analyses/
- folder for analyses of experimental and simulated data
- analyses scripts can be found in subfolder/1_srcAna/
- results are saved in subfolder/3_results/
- for some analyses intermediate data are saved in subfolder/2_dataEv/ for further evaluations
- psychExp/
- placeholder directory structure for experimental data (available at https://doi.org/10.5281/zenodo.6546572)
- simulations/
- includes the scripts to run the simulations with the model to generate all analyzed data
- script to run a single simulation (parallel.py) can be found in subfolder/1_srcSim/
- the shell scripts (run.sh) help to run multiple simulations in parallel
- data are saved in subfolder/2_dataRaw/ and subfolder/4_dataEv/, data generated for the study available at https://doi.org/10.5281/zenodo.6546572
Description of the results generated in Python. For several figures, additional image processing software was used to create the final figures (to adjust the layout and/or colors).
| Results | analysis | experimental data | simulated data | simulations | comment |
|---|---|---|---|---|---|
| Figure 3 | manuscript_global_performance/ and manuscript_global_performance_vps/ | yes | yes | 60 simulations from 014a.../ and 014b.../ each | in manuscript_global_performance/ use run.sh (which also generates further figures); in manuscript_global_performance_vps/ run make_plot.py twice, with arguments 0 and 1 |
| Figure 4 | manuscript_SRtask_results/ | yes | yes | 60 simulations from 014a.../ and 014b.../ each | use run.sh (which also generates further figures) |
| Figure 5 | manuscript_vp_learning_details/ | yes | no | - | run make_plot.py twice, with arguments 0 and 1 |
| Figure 6 | manuscript_Figure_activities_and_weightchanges/ | no | yes | 60 simulations from 014a.../ and 014b.../ and a single simulation from 007a.../ | use run.sh (which also generates further figures), various simulation times are set manually in the analysis (taken from the output file of the simulation), a change of the simulation used requires an adjustment of these times |
| Figure 7 | manuscript_weight_morphing/ | no | yes | 6400 simulations from 013.../ | |
| Figure S1 | manuscript_SRtask_results/ | yes | yes | 60 simulations from 001e.../ and 001f.../ each | use run.sh (which also generates further figures) |
| Figure S2 | manuscript_global_performance/ | no | yes | 60 simulations from 001e.../ and 001f.../ each | use run.sh (which also generates further figures) |
| Figure S3 | manuscript_global_performance/ | no | yes | 60 simulations from 014a.../ | set weight_plot=True (extra_functions.py line 267) |
| Figure S4 | manuscript_clockwise/ | yes | no | - | run make_plot.py twice, with arguments 0 and 1 |
| Figure S5 | manuscript_vp_learning_details/ | yes | no | - | run make_plot.py with argument 0 |
| Figure S6 | manuscript_SRtask_results/ | yes | yes | 60 simulations from 014c.../ and 014d.../ each | use run.sh (which also generates further figures) |
| Figure S7 | manuscript_Figure_activities_and_weightchanges/ | no | yes | 60 simulations from 001e.../ and 001f.../ | use run.sh (which also generates further figures) |
| Figure S9 | manuscript_statistics/ | yes | yes | - | see below |
All statistical tests of the study are performed in analyses/manuscript_statistics/ which requires the prior performance of the analyses: manuscript_SRtask_results/ (use run.sh), manuscript_global_performance/ (use run.sh), and manuscript_global_performance_vps/ (with arguments 0 and 1).
- GNU/Linux
- Python >= 3.7.4
- ANNarchy >= 4.7.1.4
- matplotlib >= 3.4.1
- numpy >= 1.21.3
- pandas >= 1.3.5
- pingouin >= 0.5.1
- scikit-learn >= 0.22.2.post1
- scipy >= 1.7.1
- tqdm >= 4.36.1