sgp4jax

JAX-compatible SGP4 satellite propagation.

sgp4jax is a pure-JAX implementation of the SGP4/SDP4 orbital propagator, the standard algorithm used to predict satellite positions from Two-Line Element (TLE) sets. Because the entire propagation path is written in JAX, you get:

• JIT compilation — fast repeated evaluations via jax.jit

• Automatic vectorization — propagate thousands of time steps in parallel with jax.vmap

• Automatic differentiation — compute gradients of position/velocity with respect to time (or any input) via jax.grad

Contents

• Installation
  • From PyPI
  • From source
  • Optional dependencies
• Getting Started
  • Basic propagation
  • Julian Date propagation
  • JIT compilation
  • Batch propagation with vmap
  • Gradients
  • GCRF output
  • Gravity models
  • Batch TLE parsing
  • Convenience batch functions
  • Specialised propagators
  • Batch GCRF — specialised propagators
• Examples
  • Orbit Fitting
  • 7×7 Element-Space Prior Covariance
• API Reference
  • Core Functions
  • Specialized Propagators
  • Frame Transformations
  • Covariance Transformations
  • Keplerian Propagation (no perturbations)
  • Batch GCRF Convenience Functions
  • Earth Orientation (IERS)
  • Data Structures
  • Gravity Constants