Documentation

Hironaka.Resolution.Algebraic.Tuning.Parameter

The tuning parameter and Corollary 101 at that parameter #

The library tunes with the parameter s(m) = tuningParam m = max (m − 1) 1 · lcm(2, …, m), which is admissible for [Kol07, Corollary 101] for every m ≥ 1 (it is r · lcm(2, …, m) with r = max (m − 1) 1 ≥ m − 1). Kollár takes s = m! in the proof of Theorem 103 "to avoid further choices"; m! is of the admissible form when m ≤ 2 or m ≥ 6 (by (m − 1) · lcm(2, …, m) ≤ m!, [Kol07, Aside 99.10]; Hironaka.Algebra.Local.TuningParam), whereas s(m) is admissible for every m ≥ 1 without a case distinction. The order reduction of Hironaka.Sequence uses exactly:

s(1) = max 0 1 · lcm(∅) = 1.

s(2) = max 1 1 · lcm(2) = 2.

W_{s(1)}(I) = I: with the parameter 1, W_1(I) = MC(I) = D^0(I) = I (W_one, MC_eq); the form in which the closed-embedding clauses of [Kol07, Lemma 102 (3)] and [Kol07, Theorem 103 (3)] meet the tuned ideal.

[Kol07, Corollary 101] at s = s(m) (tuning_iff_E at r = max (m − 1) 1): order reduction for (X, I, m, E) is order reduction for (X, W_{s(m)}(I), s(m), E).

[Kol07, Corollary 101] at s = s(m) (isDBalanced_W): W_{s(m)}(I) is D-balanced with respect to s(m).

[Kol07, Corollary 101] at s = s(m) (isMCInvariant_W): W_{s(m)}(I) is MC-invariant with respect to s(m).