examples

Cocofest examples

Every script below is directly runnable, e.g.:

python examples/getting_started/optimization/pulse_width_optimization.py

Each script anchors its own default asset paths (.bioMod files, etc.) to its own location on disk (Path(__file__)), so they can be run from anywhere, not just from inside their own folder.

getting_started/

Minimal, single-muscle examples. Start here.

• optimization/ — Optimal control (OCP) examples using the Ding2007 model unless noted:
  • force_tracking.py — pulse width optimization to track a target force curve.
  • frequency_optimization.py — Ding2003 model, matches a target end force with a fixed 1 Hz stimulation (no optimization on stimulation itself).
  • pulse_intensity_optimization.py — Hmed2018 model, pulse intensity optimized to match a target end force.
  • pulse_width_optimization.py — pulse width optimized to match a target end force while minimizing muscle force.
  • pulse_width_optimization_mhe.py — same problem solved with a moving time horizon (MHE), using FesMhe.
• initial_value_problem/ — Forward simulation (no optimization) via IvpFes:
  • model_integration.py — integrates DingModelFrequencyWithFatigue for 300 s to show fatigue's effect on force over time.
  • pulse_mode_example.py — compares single/doublet/triplet pulse modes on the same model.
• identification/:
  • muscle_model_id.py — simulates then identifies a Hmed2018 pulse-intensity model's parameters using OcpFesId and DataExtraction.
• multibody/ — Same pulse width/intensity optimizations as above, but driving a musculoskeletal arm model through FesMskModel:
  • pulse_width_optimization_multibody.py, pulse_intensity_optimization_multibody.py.

identification/

• force_model/ — Parameter identification (simulate then re-identify) for ding2003_model_id.py, ding2007_model_id.py, hmed2018_model_id.py.
• fatigue_model/WIP.py — placeholder; fatigue model identification is not implemented yet (work in progress, see contributing guide if you want to pick this up).

other_fes_models/

One self-contained example per FES model not already covered above: marion2009_example.py, marion2013_example.py, veltink1992_example.py (also demonstrates the combined VeltinkRienerModelPulseIntensityWithFatigue).

fes_multibody/

Full musculoskeletal, FES-driven motion examples:

• reaching/reaching_task.py — Arm26 reaching task (see README).
• elbow_flexion/ — elbow_flexion_task.py (pulse width) and frequency_optimization_multibody.py (frequency + pulse intensity).
• shoulder_abduction/ — abduction_fes_driven.py vs abduction_muscle_driven.py: same task, FES-driven vs. classic muscle-driven for comparison.
• cycling/ — hand-cycling task, the most involved example set:
  • cycling_inverse_kinematics.py — warm-starts the cycling motion with inverse kinematics/dynamics.
  • cycling_with_different_driven_methods.py — same cycling task, switchable between torque_driven/muscle_driven/fes_driven dynamics.
  • cycling_pulse_width_mhe.py — hand-cycling MHE, uses FesMheMsk (see README).
  • cycling_bayesian_mhe.py — Bayesian optimization of per-muscle cost-function weights on top of cycling_pulse_width_mhe.py, requires scikit-optimize.
  • cycling_standard_clinics.py — simulates how long a standard (non-optimized) clinical stimulation protocol can be sustained before fatigue failure.
  • physiological_weight_calculation.py, cost_functions.py — helper modules (muscle scaling, custom cost functions) imported by the scripts above, not meant to be run directly.

sensitivity/

• force_length_velocity/muscle_relationships_comparison.py — compares enabling/disabling the force-length, force-velocity and passive-force relationships (see README).

msk_models/

• model_viewer.py — live-animates a .bioMod model with slider-controlled joint angles.
• Arm26/, Seth/, Wu/ — the .bioMod musculoskeletal model files used as defaults across the examples above.