import numpy as np from skyfield import api from skyfield.constants import DAY_S, tau from skyfield.earthlib import earth_rotation_angle from skyfield.framelib import true_equator_and_equinox_of_date from skyfield.functions import A, from_spherical, length_of, mxv, rot_z from skyfield.positionlib import Geocentric, ICRF, ITRF_to_GCRS2, _GIGAPARSEC_AU from skyfield.starlib import Star from .fixes import low_precision_ERA from assay import assert_raises def test_subtraction(): p0 = ICRF((10,20,30), (40,50,60), center=0, target=499) p1 = ICRF((1,2,3), (4,5,6), center=0, target=399) p = p0 - p1 assert p.center == 399 assert p.target == 499 assert isinstance(p, Geocentric) assert tuple(p.xyz.au) == (9, 18, 27) assert tuple(p.velocity.au_per_d) == (36, 45, 54) p1.center = 1 with assert_raises(ValueError): p0 - p1 def test_separation_from_on_scalar(): p0 = ICRF((1, 0, 0)) p1 = ICRF((0, 1, 0)) assert str(p0.separation_from(p1)) == '90deg 00\' 00.0"' def test_separation_from_on_two_array_values(): p0 = ICRF(([1,1], [0,0], [0,0])) p1 = ICRF(([0,-1], [1,0], [0,0])) sep = p0.separation_from(p1) d = sep.degrees assert len(d) == 2 assert d[0] == 90.0 assert d[1] == 180.0 def test_separation_from_on_an_array_and_a_scalar(): p0 = ICRF(([1,0], [0,1], [0,0])) p1 = ICRF((0, 0, 1)) sep = p0.separation_from(p1) d = sep.degrees assert len(d) == 2 assert d[0] == 90.0 assert d[1] == 90.0 # And the other way around: sep = p1.separation_from(p0) d = sep.degrees assert len(d) == 2 assert d[0] == 90.0 assert d[1] == 90.0 def test_J2000_ecliptic_coordinates_with_and_without_a_time_array(): p0 = ICRF((1,0,0)) p1 = ICRF((0,1,0)) p2 = ICRF(((1, 0), (0, 1), (0, 0))) lat0, lon0, distance0 = p0.ecliptic_latlon(epoch=None) lat1, lon1, distance1 = p1.ecliptic_latlon(epoch=None) lat2, lon2, distance2 = p2.ecliptic_latlon(epoch=None) assert lat2.degrees[0] == lat0.degrees assert lat2.degrees[1] == lat1.degrees assert lon2.degrees[0] == lon0.degrees assert lon2.degrees[1] == lon1.degrees assert distance2.au[0] == distance0.au assert distance2.au[1] == distance1.au def test_dynamic_ecliptic_coordinates_with_and_without_a_time_array(): ts = api.load.timescale() t = ts.utc(1980) p0 = ICRF((1,0,0)) p1 = ICRF((0,1,0)) p2 = ICRF(((1, 0), (0, 1), (0, 0))) lat0, lon0, distance0 = p0.ecliptic_latlon(epoch=t) lat1, lon1, distance1 = p1.ecliptic_latlon(epoch=t) lat2, lon2, distance2 = p2.ecliptic_latlon(epoch=t) assert lat2.degrees[0] == lat0.degrees assert lat2.degrees[1] == lat1.degrees assert lon2.degrees[0] == lon0.degrees assert lon2.degrees[1] == lon1.degrees assert distance2.au[0] == distance0.au assert distance2.au[1] == distance1.au def test_frame_rotations_for_mean_of_date(): ts = api.load.timescale() t = ts.utc(2020, 11, 21) p = ICRF((1.1,1.2,1.3), t=t) lat, lon, distance1 = p.frame_latlon(true_equator_and_equinox_of_date) # Verify that the frame_latlon() coordinates match those from the # more conventional radec() call. ra, dec, distance2 = p.radec(epoch='date') assert abs(lat.arcseconds() - dec.arcseconds()) < 1e-6 assert abs(lon.arcseconds() - ra.arcseconds()) < 1e-6 assert abs(distance1.au - distance2.au) < 1e-15 # Now that we know the coordinates are good, we can use them to # rebuild a trusted x,y,z vector with which to test frame_xyz(). x1, y1, z1 = from_spherical(distance1.au, lat.radians, lon.radians) x2, y2, z2 = p.frame_xyz(true_equator_and_equinox_of_date).au assert abs(x1 - x2) < 1e-15 assert abs(y1 - y2) < 1e-15 assert abs(z1 - z2) < 1e-15 def test_position_of_radec(): epsilon = _GIGAPARSEC_AU * 1e-16 p = api.position_of_radec(0, 0) assert length_of(p.xyz.au - [_GIGAPARSEC_AU, 0, 0]) < epsilon p = api.position_of_radec(6, 0) assert length_of(p.xyz.au - [0, _GIGAPARSEC_AU, 0]) < epsilon epsilon = 2e-16 p = api.position_of_radec(12, 90, 2) assert length_of(p.xyz.au - [0, 0, 2]) < epsilon p = api.position_of_radec(12, 90, distance_au=2) assert length_of(p.xyz.au - [0, 0, 2]) < epsilon ts = api.load.timescale() epoch = ts.tt_jd(api.B1950) p = api.position_of_radec(0, 0, 1, epoch=epoch) assert length_of(p.xyz.au - [1, 0, 0]) > 1e-16 ra, dec, distance = p.radec(epoch=epoch) assert abs(ra.hours) < 1e-12 assert abs(dec.degrees) < 1e-12 assert abs(distance.au - 1) < 3e-16 def test_position_from_radec(): # Only a couple of minimal tests, since the routine is deprecated. p = api.position_from_radec(0, 0) assert length_of(p.xyz.au - [1, 0, 0]) < 1e-16 p = api.position_from_radec(6, 0) assert length_of(p.xyz.au - [0, 1, 0]) < 1e-16 def test_ssb(): ts = api.load.timescale() t = ts.utc(2025, 1, 28) p = api.SSB.at(t) z = [0,0,0] assert p.xyz.au.tolist() == z assert p.velocity.au_per_d.tolist() == z star = Star(ra_hours=12, dec_degrees=345) p.observe(star) t = ts.utc(2025, 1, [28,29]) p = api.SSB.at(t) z2 = [[0,0], [0,0], [0,0]] assert p.xyz.au.tolist() == z2 assert p.velocity.au_per_d.tolist() == z2 p.observe(star) def test_velocity_in_ITRF_to_GCRS2(): # TODO: Get test working with these vectors too, showing it works # with a non-zero velocity vector, but in that case the test will # have to be fancier in how it corrects. # r = np.array([(1, 0, 0), (1, 1 / DAY_S, 0)]).T # v = np.array([(0, 1, 0), (0, 1, 0)]).T ts = api.load.timescale() t = ts.utc(2020, 7, 17, 8, 51, [0, 1]) r = np.array([(1, 0, 0), (1, 0, 0)]).T v = np.array([(0, 0, 0), (0, 0, 0)]).T r, v = ITRF_to_GCRS2(t, r, v, True) # Rotate back to equinox-of-date before applying correction. r = mxv(t.M, r) v = mxv(t.M, v) r0, r1 = r.T v0 = v[:,0] # Apply a correction: the instantaneous velocity does not in fact # carry the position in a straight line, but in an arc around the # origin; so use trigonometry to move the destination point to where # linear motion would have carried it. angvel = (t.gast[1] - t.gast[0]) / 24.0 * tau r1 = mxv(rot_z(np.arctan(angvel) - angvel), r1) r1 *= np.sqrt(1 + angvel*angvel) actual_motion = r1 - r0 predicted_motion = v0 / DAY_S relative_error = (length_of(actual_motion - predicted_motion) / length_of(actual_motion)) acceptable_error = 1e-11 assert relative_error < acceptable_error def test_light_time_method(): p = ICRF([0.0, 1.0, 0.0]) assert abs(p.light_time - 0.0057755183) < 1e-10 def test_hadec(): # If the DE430 ephemeris excerpt is avaiable, this test can run # locally against the HA number from first line of # `moon_topo_4_6_2017_mkb_sf_v5_hadec.csv.txt` at: # https://github.com/skyfielders/python-skyfield/issues/510 #planets = api.load('de430_1850-2150.bsp') #expected_ha = -0.660078756021 # But in CI, we use DE421 for space and speed. planets = api.load('de421.bsp') expected_ha = -0.660078752 ts = api.load.timescale() ts.polar_motion_table = [0.0], [0.009587], [0.384548] t = ts.utc(2017, 4, 6) topos = api.wgs84.latlon(-22.959748, -67.787260, elevation_m=5186.0) earth = planets['Earth'] moon = planets['Moon'] a = (earth + topos).at(t).observe(moon).apparent() ha, dec, distance = a.hadec() difference_mas = (ha.hours - expected_ha) * 15 * 3600 * 1e3 assert abs(difference_mas) < 0.03 # Drive-by test of position repr. assert repr(a) == ( '' ) # Test that the CIRS coordinate of the TIO is consistent with the Earth Rotation Angle # This is mostly an internal consistency check def test_cirs_era(): ts = api.load.timescale() st = ts.utc(year=np.arange(1951, 2051)) planets = api.load('de421.bsp') pos = planets['earth'] + api.Topos(longitude_degrees=0.0, latitude_degrees=0.0) # Get the TIO tio = pos.at(st).from_altaz(alt_degrees=90, az_degrees=180) # Get the TIOs RA in CIRS coordinates, and the Earth Rotation Angle tio_ra, tio_dec, _ = tio.cirs_radec(st) era = 360.0 * earth_rotation_angle(st.ut1) tol = (1e-8 / 3600.0) # 10 nano arc-second precision assert np.allclose(tio_ra.degrees, era, rtol=0.0, atol=tol) assert np.allclose(tio_dec.degrees, 0.0, rtol=0.0, atol=tol) # Check a line of points along the terrestrial prime meridian all have the same # CIRS RA, and that their declinations are correct. def test_cirs_meridian(): ts = api.load.timescale() st = ts.utc(year=2051) planets = api.load('de421.bsp') pos = planets['earth'] + api.Topos(longitude_degrees=0.0, latitude_degrees=0.0) # Get a series of points along the meridian alt = np.arange(1, 90) meridian = pos.at(st).from_altaz(alt_degrees=alt, az_degrees=0.0) # Get the TIOs RA in CIRS coordinates, and the Earth Rotation Angle md_ra, md_dec, _ = meridian.cirs_radec(st) era = 360.0 * earth_rotation_angle(st.ut1) tol = (1e-7 / 3600.0) # 100 nano arc-second precision assert np.allclose(md_ra.degrees, era, rtol=0.0, atol=tol) assert np.allclose(md_dec.degrees, 90 - alt, rtol=0.0, atol=tol) # Check a set of positions and times against results calculated externally # using the IAU SOFA library (20180130 release). For reference the code used # was: # # #include # #include # #include # # int main(int argc, char ** argv) { # # // Test data as RA, DEC, TDB. Positions ICRS(deg), time in JD. # double test_data[3][3] = { # {45.0, 46.0, 2458327}, # {200.0, -22.0, 2458327}, # {45.0, 46.0, 2459327} # }; # # for(int i = 0; i < 3; i++) { # double ra_icrs = test_data[i][0] / 180.0 * M_PI; # double dec_icrs = test_data[i][1] / 180.0 * M_PI; # double jd_tdb = test_data[i][2]; # # double ra_cirs, dec_cirs; # double eo; # # iauAtci13(ra_icrs, dec_icrs, 0.0, 0.0, 0.0, 0.0, jd_tdb, 0.0, # &ra_cirs, &dec_cirs, &eo); # # printf("%.12f %.12f\n", ra_cirs / (2 * M_PI) * 360, # dec_cirs / (2 * M_PI) * 360); # } # } def test_cirs_sofa(): ts = api.load.timescale() earth = api.load('de421.bsp')['earth'] test_data = [ [45.0, 46.0, 2458327], [200.0, -22.0, 2458327], [45.0, 46.0, 2459327] ] # Results output by SOFA. Calculated using the source code above. sofa_results = [ [45.074343838325, 46.067831092355], [200.013551320030, -22.096008994214], [45.077698288877, 46.082296559677] ] tol = 1e-5 / 3600.0 # 10 micro arc-seconds for ((ra_icrs, dec_icrs, tdb), (ra_sofa, dec_sofa)) in zip(test_data, sofa_results): ss = Star(ra_hours=(ra_icrs / 15.0), dec_degrees=dec_icrs) st = ts.tdb(jd=tdb) with low_precision_ERA(): ra_cirs, dec_cirs, _ = earth.at(st).observe(ss).apparent().cirs_radec(st) assert np.allclose(ra_cirs.degrees, ra_sofa, rtol=0.0, atol=tol) assert np.allclose(dec_cirs.degrees, dec_sofa, rtol=0.0, atol=tol) def test_phase_angle_and_fraction_illuminated(): ts = api.load.timescale() t = ts.utc(2018, 9, range(9, 19), 5) t1 = t[-1] e = api.load('de421.bsp') earth, moon, sun = e['earth'], e['moon'], e['sun'] p = earth.at(t1).observe(moon) a = p.phase_angle(sun).degrees.round(1) assert a == 76.2 i = p.fraction_illuminated(sun).round(2) assert i == 0.62 p = earth.at(t).observe(moon) a = p.phase_angle(sun).degrees.round(1) assert list(a) == [172.0, 172.6, 159.6, 146.6, 133.9, 121.7, 109.9, 98.4, 87.2, 76.2] i = (p.fraction_illuminated(sun) * 100).round(2) assert list(i) == [0.49, 0.41, 3.12, 8.25, 15.30, 23.73, 33.01, # not 33.02? 42.71, 52.45, 61.92] def test_astropy_conversion(): try: import astropy except ImportError: # Drat: assay doesn't know about skipping a test. #raise SkipTest('AstroPy not installed') return else: astropy # Use the library's name, to avoid a linter complaint. ts = api.load.timescale() r = np.array([1, 2, 3]) t = ts.tt(2022, 1, 3) p = ICRF(r, t=t, center=0) a = p.to_skycoord() assert str(a) == '' assert a.obstime is None p = ICRF(r, t=t, center=399) a = p.to_skycoord() assert str(a) == ( '' ) assert a.obstime.fits == '2022-01-03T00:00:00.000' p = ICRF(r, t=t, center=3) with assert_raises(NotImplementedError): a = p.to_skycoord() def test_old_position_attribute(): ts = api.load.timescale() t = ts.tt(2022, 1, 3) r = A[6, 7, 8] p = ICRF(r, t=t, center=0) assert tuple(p.position.au) == (6, 7, 8)