The discriminant identity (τ − tauConj)² = d #
Thesis. §3.2, the discriminant of the order: for an honest discriminant
d (i.e. d ≡ 0, 1 (mod 4)), (τ − tauConj)² = d.
This file proves:
tau_sub_tauConj—τ − tauConj = 2τ − dtau_sub_tauConj_sq— the general identity, valid for alldtau_sub_tauConj_sq_of_valid_disc— specialises to(τ − tauConj)² = dwhend ≡ 0 ∨ 1 (mod 4)(the case used in the thesis)
Divergence from thesis. The thesis only ever needs (τ − tauConj)² = d because
it assumes d is a discriminant. Because QuadraticOrder d is defined for all
d here (see Basic.lean), the honest general identity carries the Euclidean
remainder term: (τ − tauConj)² = (d² − 4⌊(d²−d)/4⌋) • 1, which collapses to d
exactly under the congruence hypothesis. The general form
tau_sub_tauConj_sq is a Lean-only artifact with no thesis counterpart.
Discriminant: (τ - tauConj)² #
General discriminant identity: (τ - tauConj)² = (d² - 4·⌊(d²-d)/4⌋) • 1. The
right-hand-side equals d • 1 when d ≡ 0 or d ≡ 1 (mod 4) (the values
for which poly d has integer coefficients matching the discriminant of the
quadratic field). For general d, the difference is the Euclidean remainder
(d² - d) mod 4.
Under the valid-discriminant hypothesis d ≡ 0 ∨ d ≡ 1 (mod 4), the
discriminant equals d exactly: (τ - tauConj)² = d • 1. This is the case used
throughout the thesis (the only d for which QuadraticOrder d coincides
with the quadratic order of discriminant d).