PhiFlow
PlaygroundBeginner Guides
Installation Instructions: Requirements, installation, CUDA compilation
Cookbook: Example code snippets for various things
Tensors and Dimensions
Differentiation
Fluids
- ▶️ Introduction Video
- Differentiable fluid simulations
- Higher-order incompressible fluids
- Batched Obstacles
I/O
Visualization
Neural Networks
Advanced
- What to Avoid: How to keep your code compatible with PyTorch, TensorFlow and Jax
Module Documentation
| Module API | Documentation |
|---|---|
| phi.vis | Visualization: Plotting, interactive user interfaces Dash: Web interface Console: Command line interface |
| phi.physics phi.physics.advect phi.physics.fluid phi.physics.diffuse |
Fluids Tutorial: Introduction to core classes and fluid-related functions. Higher-order schemes: Compares the accuracy of various numerial schemes. Overview: Domains, built-in physics functions Functions for Fluid Simulations: Advection, projection, diffusion |
| phi.field | Overview: Grids, particles Staggered Grids: Data layout, usage Reading and Writing Simulation Data Scene Format Specification: Directory layout, file format |
| phi.geom | Overview: Differentiable Geometry |
| phi.math phi.math.backend phi.math.extrapolation phi.math.magic |
Overview: Named dimensions, backends, indexing, non-uniform tensors, precision Optimization and Training: Automatic differentiation, neural network training Performance: GPU, JIT compilation, profiler |
| phi.torch.nets phi.tf.nets phi.jax.stax.nets |
Built-in Neural Networks: Architectures, convenience functions |
Core Classes
API Documentation
The API documentation is generated using pdoc3. To manually generate the documentation, add the PhiFlow directory to your Python path and run
$ pdoc --html --output-dir docs --force phi
This requires PyTorch, TensorFlow and Jax to be installed, in addition to the standard ΦFlow requirements.
Contributing to ΦFlow
Contributions are welcome!
If you have changes to merge, check out our style guide before opening a pull request.