def _true_ecliptic_rotation_matrix(equinox):
# This code calls the same routines as done in pnm06a from ERFA, which
# retrieves the precession matrix (including frame bias) according to
# the IAU 2006 model, and including the nutation.
# This family of systems is less popular
# (see https://github.com/astropy/astropy/pull/6508).
jd1, jd2 = get_jd12(equinox, 'tt')
# Here, we call the three routines from erfa.pnm06a separately,
# so that we can keep the nutation for calculating the true obliquity
# (which is a fairly expensive operation); see gh-11000.
# pnm06a: Fukushima-Williams angles for frame bias and precession.
# (ERFA names short for F-W's gamma_bar, phi_bar, psi_bar and epsilon_A).
gamb, phib, psib, epsa = erfa.pfw06(jd1, jd2)
# pnm06a: Nutation components (in longitude and obliquity).
dpsi, deps = erfa.nut06a(jd1, jd2)
# pnm06a: Equinox based nutation x precession x bias matrix.
rnpb = erfa.fw2m(gamb, phib, psib + dpsi, epsa + deps)
# calculate the true obliquity of the ecliptic
obl = erfa.obl06(jd1, jd2) + deps
return matrix_product(rotation_matrix(obl << u.radian, 'x'), rnpb)
def gcrs_precession_mat(equinox):
gamb, phib, psib, epsa = erfa.pfw06(*get_jd12(equinox, 'tt'))
return erfa.fw2m(gamb, phib, psib, epsa)