The Skyfield library provides a simple and safe interface to high-precision astronomical routines that are written in pure Python, with NumPy as their only binary dependency; this library itself requires no compilation.
You can install the current development version of Skyfield with:
pip install https://github.com/brandon-rhodes/python-skyfield/archive/master.zipThe interface is still evolving, but you can already ask the position of a planet in the sky using one of three simple techniques:
import skyfield.coordinates
import skyfield.planets
import de421
ephemeris = skyfield.planets.Ephemeris(de421)
earth = ephemeris.earth
mars = ephemeris.mars
# Getting the Julian date is not yet convenient.
from sgp4.ext import jday
t = jday(2013, 1, 30, 16, 8, 23) # Julian date
# What is the sky-catalog coordinate of Mars?
radec = mars.observe_from(earth(t)).astrometric()
# Where is Mars in the sky, viewed from Earth's center?
radec = mars.observe_from(earth(t)).apparent()
# Where is Mars in the sky, viewed from Boston?
boston = coordinates.Topos('71.0603 W', '42.3583 N',
elevation=0.0, temperature=10.0, pressure=1010.0)
radec = mars.observe_from(boston(t)).apparent()All of these routines are designed to operate very efficiently if, instead of providing them with a single Julian date, you provide them with a NumPy array of dates instead:
import numpy as np
t0 = jday(2013, 1, 30, 0, 0, 0)
t = np.arange(t0, t0 + 365.0, 1.0)
radec = mars.observe_from(boston(t)).apparent()I will expand this README as the API continues to develop, but I wanted to at least get these few details typed up to help programmers who are already interested in trying out Skyfield!