This repository is associated to the following publication (under review):
Q. Vacher, N. Beuve, P. Allaire, T. Marty, M. Dardaillon and K. Desnos. Hybrid Genetic Programming and Deep Reinforcement Learning for Low-Complexity Robot Arm Trajectory Planning
The repository contains:
- Code and scripts to reproduce the experiments presented in the paper.
- Experimental data and logs produced by the authors and presented in the paper.
├─ ECTA24-artifacts # root folder
│ │
│ ├─ gegelati # git submodule pointing to gegelati develop commit: 9b4092f
│ │ │...
│ │
│ ├─ armlearn-wrapper # git submodule pointing to the ArmLearn Environment
│ │ │... # wrapper for gegelati.
│ │
│ ├─ mainAnalysis.ipynb # Jupyter Notebook with the main different plots and analysis
│ │
│ ├─ moreLessValuableCurves.ipynb # Jupyter Notebook with other different plots and analysis
│ │
│ ├─ data # Experimental data.
│ │ │
│ │ │
│ │ │
│ │ ├─ experimentalStudy1 # Folder containing the performances of the first experimental study
│ │ │ │
│ │ │ ├─ multiTraining_x # One configuration with params and different seeds.
│ │ │ │
│ │ │ │
│ │ │ │ ├─ params # Parameter forlder containing the different parameters
│ │ │ │ │
│ │ │ │ │
│ │ │ │ ├─ outLogs # Folder of logs.
│ │ │ │ │
│ │ │ │ │
│ │ │ │ │ ├─ logsGegelati.ods # Logs from the training.
│ │ │ │ │ │
│ │ │ │ │ ├─ bestPolicyStats.md # Statistics of champion TPGs throughout the generations.
│ │ │ │ │ │
│ │ │ │ │ ├─ out_best.dot # Champion TPG after all generations.
│ │ │ │ │ │
│ │ │ │ │ ├─ out_best_stats.md # Statistics about the champion.
│ │ │ │ │ │
│ │ │ │ │ ├─ outputGegelati.csv # Logs from the testing.
│ │ │ ├─ ...
│ │ │
│ │ ├─ expeStudy2 # Folder containing the performances of the second experimental study
│ │ │
│ │ ├─ ... # All the other data for SAC and hybrid solution, in the same format
│ │
│ ├─ results # Folder with the different figures obained