Getting Started with Firedrake
This section covers the standard reinforcement learning interface for HydroGym's Firedrake-based flow environments, using the familiar env.reset() and env.step() API from Gymnasium.
Cavity
The scripts on this page access the Firedrake solver directly — they do not use the standard env.reset() / env.step() interface. For RL training, see Getting Started.
Cylinder
The scripts on this page access the Firedrake solver directly — they do not use the standard env.reset() / env.step() interface. For RL training, see Getting Started.
Pinball
The scripts on this page access the Firedrake solver directly — they do not use the standard env.reset() / env.step() interface. For RL training, see Getting Started.
Backward-Facing Step
The scripts on this page access the Firedrake solver directly — they do not use the standard env.reset() / env.step() interface. For RL training, see Getting Started.