Read an installed native incumbent from its stable code array and checked label. This semantic observation does not expand sparse rows.
Equations
- One or more equations did not get rendered due to their size.
Instances For
def
Hex.GraphIso.Nauty.Sparse.State.key
{n : Nat}
(G : SparseGraph n)
(bs : List Nat)
(st : State n)
:
During code overwriting, the semantic incumbent retains its complete saved code sequence. Its graph is still the actual saved native label.
Equations
- One or more equations did not get rendered due to their size.
Instances For
theorem
Hex.GraphIso.Nauty.Sparse.best_eq_key
{n : Nat}
{G : SparseGraph n}
{cs bs : List Nat}
{st : State n}
{comparison : Int}
(h : CodeCmpInv n cs bs st.canoncode st.canonlevel st.eqlevCanon comparison)
(hne : comparison ≠ 1)
:
A stable canonical code array reads exactly the semantic sparse key.
theorem
Hex.GraphIso.Nauty.Sparse.settled_read
{n : Nat}
{G : SparseGraph n}
{cs bs : List Nat}
{st : State n}
(h : Settled cs bs st)
:
Either settled leaf verdict exposes the same native incumbent.
theorem
Hex.GraphIso.Nauty.Sparse.firstterminal_best
{n : Nat}
{G : SparseGraph n}
{cs : List Nat}
{st : State n}
{l : Label n}
(h : Codes cs cs (firstterminal cs.length st))
(hne : cs ≠ [])
(hl : Label.ofArray? n st.lab = some l)
:
State.best G (firstterminal cs.length st) = some { codes := cs ++ [codeSentinel], graph := G.relabel l.perm }
Installing the first leaf exposes precisely its executed code chain and sparse relabelling, once the code machine has been initialized.