coordinax

Coordinates in JAX

Coordinax enables calculations with coordinates in JAX. Built on Equinox and Quax.

Installation

pip install coordinax

Documentation

Coming soon. In the meantime, if you've used astropy.coordinates, then coordinax should be fairly intuitive.

Quick example

import coordinax as cx
import jax.numpy as jnp
from unxt import Quantity

q = cx.CartesianPos3D(
    x=Quantity(jnp.arange(0, 10.0), "km"),
    y=Quantity(jnp.arange(5, 15.0), "km"),
    z=Quantity(jnp.arange(10, 20.0), "km"),
)
print(q)
# <CartesianPos3D (x[km], y[km], z[km])
#     [[ 0.  5. 10.]
#      [ 1.  6. 11.]
#      ...
#      [ 8. 13. 18.]
#      [ 9. 14. 19.]]>

q2 = cx.represent_as(q, cx.SphericalPos)
print(q2)
# <SphericalPos (r[km], theta[rad], phi[rad])
#     [[11.18   0.464  1.571]
#      [12.57   0.505  1.406]
#      ...
#      [23.601  0.703  1.019]
#      [25.259  0.719  0.999]]>

p = cx.CartesianVel3D(
    d_x=Quantity(jnp.arange(0, 10.0), "m/s"),
    d_y=Quantity(jnp.arange(5, 15.0), "m/s"),
    d_z=Quantity(jnp.arange(10, 20.0), "m/s"),
)
print(p)
# <CartesianVel3D (d_x[m / s], d_y[m / s], d_z[m / s])
#     [[ 0.  5. 10.]
#      [ 1.  6. 11.]
#      ...
#      [ 8. 13. 18.]
#      [ 9. 14. 19.]]>

p2 = cx.represent_as(p, cx.SphericalVel, q)
print(p2)
# <SphericalVel (d_r[m / s], d_theta[m rad / (km s)], d_phi[m rad / (km s)])
#     [[ 1.118e+01 -3.886e-16  0.000e+00]
#      [ 1.257e+01 -1.110e-16  0.000e+00]
#      ...
#      [ 2.360e+01  0.000e+00  0.000e+00]
#      [ 2.526e+01 -2.776e-16  0.000e+00]]>

Citation

If you found this library to be useful in academic work, then please cite.

Development

We welcome contributions!