Skip to content

data.md

Latest commit

 

History

History
49 lines (40 loc) · 2.1 KB

data.md

File metadata and controls

49 lines (40 loc) · 2.1 KB

Data Directory Structure

data/ — Configuration & Resources (~105 MB, in repo)

Directory Description
gait_generator/ RL training metadata YAML configs (10 experiments)
surrogate/ Surrogate model training configs (regression, GP)
record/ Rollout configs defining which signals to export
human/ Skeleton XML definitions
muscle/ Hill-type muscle model parameters
device/ Exoskeleton device XMLs (hip GEMS2)
OBJ/ 3D meshes for OpenGL visualization
figs/ Pre-generated figure images for plot scripts
res/ Runtime resources

experiment_data/ — Experiment Data

All pipeline outputs used in the paper (~20 GB). Download separately from onedrive and unzip at the project root. Required for figure reproduction and for resuming the pipeline from intermediate steps without re-running training.

Directory Description
checkpoints/ RL policy checkpoints (Step 1 output)
params/ Parameter sweep CSVs/parquets
simulation/ Simulation rollout outputs (Step 2 output)
tfevents/ TensorBoard training logs
real/ Real-world reference gait data (GEMS, EMG, Gutenberg)

Pipeline data flow

data/gait_generator/*.yaml        --> Step 1: RL training configs
experiment_data/checkpoints/      --> Step 2: TorchScript conversion input
experiment_data/params/*.csv      --> Step 2: rollout input parameters
data/record/*.yaml                --> Step 2: rollout recording configs
experiment_data/simulation/       --> Step 3: surrogate training input
data/surrogate/*.yaml             --> Step 3: surrogate training configs
experiment_data/real/             --> figure reproduction
experiment_data/tfevents/         --> figure reproduction

Setup

  1. Clone the repository (includes data/).
  2. Obtain experiment_data.zip separately.
  3. Unzip at the project root: 
    unzip experiment_data.zip -d project-exo-plore/