:: Chordal Graphs
:: by Broderick Arneson and Piotr Rudnicki
::
:: Received August 18, 2006
:: Copyright (c) 2006-2016 Association of Mizar Users
:: (Stowarzyszenie Uzytkownikow Mizara, Bialystok, Poland).
:: This code can be distributed under the GNU General Public Licence
:: version 3.0 or later, or the Creative Commons Attribution-ShareAlike
:: License version 3.0 or later, subject to the binding interpretation
:: detailed in file COPYING.interpretation.
:: See COPYING.GPL and COPYING.CC-BY-SA for the full text of these
:: licenses, or see http://www.gnu.org/licenses/gpl.html and
:: http://creativecommons.org/licenses/by-sa/3.0/.
environ
vocabularies NUMBERS, XCMPLX_0, ORDINAL1, ARYTM_1, XXREAL_0, NAT_1, CARD_1,
ARYTM_3, ABIAN, SUBSET_1, RELAT_1, INT_1, FINSEQ_1, FUNCT_1, FINSEQ_4,
XBOOLE_0, FINSET_1, GRAPH_2, ORDINAL4, GLIB_000, GLIB_001, TARSKI,
ZFMISC_1, RCOMP_1, GRAPH_1, RELAT_2, REWRITE1, FUNCOP_1, GLIB_002,
PARTFUN1, MEMBERED, TOPGEN_1, CHORD;
notations TARSKI, XBOOLE_0, SUBSET_1, XXREAL_2, ORDINAL1, INT_1, XCMPLX_0,
XXREAL_0, DOMAIN_1, RELAT_1, FUNCT_1, FUNCT_2, FINSEQ_1, CARD_1,
FINSET_1, NAT_1, ZFMISC_1, GLIB_000, GLIB_001, GLIB_002, FUNCOP_1, ABIAN,
ENUMSET1, FINSEQ_4, NUMBERS, GRAPH_2, MEMBERED;
constructors DOMAIN_1, REAL_1, NAT_D, FINSEQ_4, GRAPH_2, GLIB_001, GLIB_002,
VALUED_1, XXREAL_2, RELSET_1, NUMBERS;
registrations XBOOLE_0, SUBSET_1, RELAT_1, FUNCT_1, ORDINAL1, FUNCOP_1,
FINSET_1, NUMBERS, XXREAL_0, XREAL_0, NAT_1, INT_1, MEMBERED, FINSEQ_1,
GLIB_000, ABIAN, GRAPH_2, GLIB_001, GLIB_002, FUNCT_2, XXREAL_2, CARD_1,
RELSET_1;
requirements ARITHM, BOOLE, NUMERALS, REAL, SUBSET;
begin :: Preliminaries
theorem :: CHORD:1
for n being non zero Nat holds n-1 is Nat & 1 <= n;
theorem :: CHORD:2 :: Nat02
for n being odd Nat holds n-1 is Nat & 1 <= n;
theorem :: CHORD:3 :: EvenOdd02
for n,m being odd Integer st n < m holds n <= m-2;
theorem :: CHORD:4 :: EvenOdd03:
for n,m being odd Integer st m < n holds m+2 <= n;
theorem :: CHORD:5 :: EvenOdd04:
for n being odd Nat st 1 <> n
ex m being odd Nat st m+2 = n;
theorem :: CHORD:6 :: Odd100
for n being odd Nat st n<=2 holds n=1;
theorem :: CHORD:7 :: Odd101
for n being odd Nat st n<=4 holds n=1 or n=3;
theorem :: CHORD:8 :: Odd102
for n being odd Nat st n<=6 holds n=1 or n=3 or n=5;
theorem :: CHORD:9
for n being odd Nat st n<=8 holds n=1 or n=3 or n=5 or n=7;
theorem :: CHORD:10 :: Even100
for n being even Nat st n<=1 holds n=0;
theorem :: CHORD:11 :: Even101
for n being even Nat st n<=3 holds n=0 or n=2;
theorem :: CHORD:12 :: Even102
for n being even Nat st n<=5 holds n=0 or n=2 or n=4;
theorem :: CHORD:13 :: Even103
for n being even Nat st n<=7 holds n=0 or n=2 or n=4 or n=6;
theorem :: CHORD:14
for p being FinSequence, n being non zero Nat st p is
one-to-one & n <= len p holds (p.n)..p = n;
theorem :: CHORD:15 :: Index01
for p being non empty FinSequence, T being non empty Subset of
rng p ex x being set st x in T & for y being set st y in T holds x..p <= y..p
;
definition
let p be FinSequence, n be Nat;
func p.followSet(n) -> finite set equals
:: CHORD:def 1
rng (n,len p)-cut p;
end;
theorem :: CHORD:16 :: Follow00
for p being FinSequence, x being set, n being Nat st
x in rng p & n in dom p & p is one-to-one holds x in p.followSet(n) iff x..p >=
n;
theorem :: CHORD:17 :: Follow03
for p, q being FinSequence, x being set st p = <*x*>^q for n
being non zero Nat holds p.followSet(n+1) = q.followSet(n);
theorem :: CHORD:18 :: FinSubseq00
for X being set, f being FinSequence of X, g being Subset of f
st len Seq g = len f holds Seq g = f;
begin :: Miscellany on graphs
theorem :: CHORD:19
for G being _Graph, S being Subset of the_Vertices_of G for H
being inducedSubgraph of G,S
for u,v being object st u in S & v in S
for e being object st e Joins u,v,G holds e Joins u,v,H;
theorem :: CHORD:20
for G being _Graph, W being Walk of G holds W is Trail-like iff len W
= 2*(card W.edges())+1;
theorem :: CHORD:21 :: Walk02
for G being _Graph, S being Subset of the_Vertices_of G for H
being removeVertices of G,S for W being Walk of G st
(for n being odd Nat st n <= len W holds not W.n in S)
holds W is Walk of H;
theorem :: CHORD:22 :: Walk03
for G being _Graph, a,b be set st a<>b for W being Walk of G st
W.vertices() = {a,b} holds ex e being set st e Joins a,b,G;
theorem :: CHORD:23 :: Walk04
for G being _Graph, S being non empty Subset of the_Vertices_of
G for H being inducedSubgraph of G,S for W being Walk of G st W.vertices() c= S
holds W is Walk of H;
theorem :: CHORD:24 :: Cyclelike01
for G1,G2 being _Graph st G1 == G2 for W1 be Walk of G1, W2
being Walk of G2 st W1 = W2 holds W1 is Cycle-like implies W2 is Cycle-like;
theorem :: CHORD:25 :: Path01
for G being _Graph, P being Path of G, m, n being odd Nat
st m <= len P & n <= len P & P.m = P.n holds m = n or m = 1 & n = len P
or m = len P & n = 1;
theorem :: CHORD:26
for G being _Graph, P being Path of G st P is open for a,e,b being set
st not a in P.vertices() & b = P.first() & e Joins a,b,G holds G.walkOf(a,e,b)
.append(P) is Path-like;
theorem :: CHORD:27 :: PathLike03
:: A similar theorem is needed for obtaining non closed Path
for G being _Graph, P,H being Path of G st P.edges() misses H
.edges() & P is open & H is non trivial & H is open & P.vertices() /\ H
.vertices() = { P.first(), P.last() } & H.first() = P.last() & H.last() = P
.first() holds P.append(H) is Cycle-like;
theorem :: CHORD:28
for G being _Graph, W1,W2 being Walk of G st W1.last() = W2
.first() holds W1.append(W2).length() = W1.length() + W2.length();
theorem :: CHORD:29 :: Subgraph01
for G being _Graph, A,B being non empty Subset of
the_Vertices_of G st B c= A for H1 being inducedSubgraph of G,A for H2 being
inducedSubgraph of H1,B holds H2 is inducedSubgraph of G,B;
theorem :: CHORD:30 :: Subgraph01a
for G being _Graph, A,B being non empty Subset of
the_Vertices_of G st B c= A for H1 being inducedSubgraph of G,A for H2 being
inducedSubgraph of G,B holds H2 is inducedSubgraph of H1,B;
theorem :: CHORD:31 :: Subgraph02
for G being _Graph, S,T being non empty Subset of
the_Vertices_of G st T c= S for G2 being inducedSubgraph of G,S holds G2
.edgesBetween(T) = G.edgesBetween(T);
:: we do not consider infinite graphs
scheme :: CHORD:sch 1
FinGraphOrderCompInd{P[set]}: for G being finite _Graph holds P[G]
provided
for k being non zero Nat st (for Gk being finite _Graph
st Gk.order() < k holds P[Gk]) holds for Gk1 being finite _Graph st Gk1.order()
= k holds P[Gk1];
:: should be in GLIBs
theorem :: CHORD:32
for G being _Graph, W being Walk of G st W is open & W is
Path-like holds W is vertex-distinct;
:: PathLike15
:: should be in GLIB
theorem :: CHORD:33
for G being _Graph, P being Path of G st P is open & len P > 3
for e being object st e Joins P.last(),P.first(),G holds P.addEdge(e) is
Cycle-like;
begin :: Shortest topological path
definition
let G be _Graph, W be Walk of G;
attr W is minlength means
:: CHORD:def 2
for W2 being Walk of G st W2 is_Walk_from W
.first(),W.last() holds len W2 >= len W;
end;
theorem :: CHORD:34 :: WalkSubwalk00
for G being _Graph, W being Walk of G, S being Subwalk of W st S
.first() = W.first() & S.edgeSeq() = W.edgeSeq() holds S = W;
theorem :: CHORD:35 :: LenSubwalk00
for G being _Graph, W being Walk of G, S being Subwalk of W st
len S = len W holds S = W;
theorem :: CHORD:36
for G being _Graph, W being Walk of G st W is minlength holds W is Path-like;
::$CT
theorem :: CHORD:38
for G being _Graph, W being Walk of G holds (for P being Path of
G st P is_Walk_from W.first(),W.last() holds len P >= len W) implies W is
minlength;
theorem :: CHORD:39 :: TopPath02
for G being _Graph, W being Walk of G ex P being Path of G st P
is_Walk_from W.first(),W.last() & P is minlength;
theorem :: CHORD:40 :: TopPath03
for G being _Graph, W being Walk of G st W is minlength holds
for m,n being odd Nat st m+2 < n & n <= len W holds
not ex e being object st e Joins W.m,W.n,G;
theorem :: CHORD:41 :: TopPath04
for G being _Graph, S being non empty Subset of the_Vertices_of
G for H being inducedSubgraph of G,S for W being Walk of H st W is minlength
for m,n being odd Nat st m+2 < n & n <= len W holds
not ex e being object st e Joins W.m,W.n,G;
theorem :: CHORD:42 :: TopPath05
for G being _Graph for W being Walk of G st W is minlength
for m,n being odd Nat st m<=n & n<=len W holds
W.cut(m,n) is minlength;
theorem :: CHORD:43
for G being _Graph st G is connected for A,B being non empty Subset of
the_Vertices_of G st A misses B holds ex P being Path of G st P is minlength &
P is non trivial & P.first() in A & P.last() in B & for n being odd Nat
st 1 < n & n < len P holds not P.n in A & not P.n in B;
begin :: Adjacency and complete graphs
definition
let G be _Graph, a,b be Vertex of G;
pred a,b are_adjacent means
:: CHORD:def 3 :: DefAdjacent
ex e being object st e Joins a,b,G;
symmetry;
end;
theorem :: CHORD:44 :: VAdjacent00
for G1,G2 being _Graph st G1 == G2 for u1,v1 being Vertex of G1
st u1,v1 are_adjacent for u2,v2 being Vertex of G2 st u1=u2 & v1=v2 holds u2,v2
are_adjacent;
theorem :: CHORD:45 :: VAdjacent01
for G being _Graph, S being non empty Subset of the_Vertices_of
G for H being inducedSubgraph of G,S for u, v being Vertex of G, t, w being
Vertex of H st u = t & v = w holds u,v are_adjacent iff t,w are_adjacent;
theorem :: CHORD:46 :: PathLike05
for G being _Graph, W being Walk of G st W.first() <> W.last() &
not W.first(),W.last() are_adjacent holds W.length() >= 2;
theorem :: CHORD:47 :: PathBuilder00
:: add sequences of vertices and edges:: PathBuilder01
for G being _Graph, v1,v2,v3 being Vertex of G st v1<>v2 & v1<>
v3 & v2<>v3 & v1,v2 are_adjacent & v2,v3 are_adjacent
ex P being Path of G, e1,e2 being object
st P is open & len P = 5 & P.length() = 2 & e1 Joins v1,v2,G & e2
Joins v2,v3,G & P.edges() = {e1,e2} & P.vertices() = {v1,v2,v3} & P.1 = v1 & P.
3 = v2 & P.5 = v3;
theorem :: CHORD:48
for G being _Graph, v1,v2,v3,v4 being Vertex of G st v1<>v2 & v1
<>v3 & v2<>v3 & v2<>v4 & v3<>v4 & v1,v2 are_adjacent & v2,v3 are_adjacent & v3,
v4 are_adjacent ex P being Path of G st len P = 7 & P.length() = 3 & P
.vertices() = {v1,v2,v3,v4} & P.1 = v1 & P.3 = v2 & P.5 = v3 & P.7 = v4;
definition
let G be _Graph, S be set;
func G.AdjacentSet(S) -> Subset of the_Vertices_of G equals
:: CHORD:def 4
{u where u is
Vertex of G : not u in S & ex v being Vertex of G st (v in S & u,v are_adjacent
)};
end;
theorem :: CHORD:49
for G being _Graph, S, x being set st x in G.AdjacentSet(S) holds not x in S;
theorem :: CHORD:50 :: Adjacent00
for G being _Graph, S being set for u being Vertex of G holds u
in G.AdjacentSet(S) iff not u in S & ex v being Vertex of G st v in S & u,v
are_adjacent;
theorem :: CHORD:51 :: Adjacent01
for G1,G2 being _Graph st G1 == G2 for S being set holds G1
.AdjacentSet(S) = G2.AdjacentSet(S);
theorem :: CHORD:52 :: AdjacentV00
for G being _Graph, u,v being Vertex of G holds u in G
.AdjacentSet({v}) iff u <> v & v,u are_adjacent;
theorem :: CHORD:53
for G being _Graph, x,y being set holds x in G.AdjacentSet({y}) iff y
in G.AdjacentSet({x});
theorem :: CHORD:54
for G being _Graph, C being Path of G st C is Cycle-like & C.length()
> 3 for x being Vertex of G st x in C.vertices() ex m,n being odd Nat st m+2 <
n & n <= len C & not (m=1 & n = len C) & not (m=1 & n = len C-2) & not (m=3 & n
= len C) & C.m <> C.n & C.m in G.AdjacentSet({x}) & C.n in G.AdjacentSet({x})
;
theorem :: CHORD:55 :: Cycle01a
for G being _Graph, C being Path of G st C is Cycle-like & C
.length() > 3 for x being Vertex of G st x in C.vertices() ex m,n being odd
Nat st m+2 < n & n <= len C & C.m <> C.n &
C.m in G.AdjacentSet({x})
& C.n in G.AdjacentSet({x}) &
for e being object st e in C.edges() holds not e Joins C.m,C.n,G;
:: Gilbert's definition of isolated does not allow a vertex to have a
:: loop and a vertex just with a loop on it is NOT isolated.
:: This needs to be fixed, e.g.
:: v is isolated means G.AdjacentSet({v}) = {}
:: But we can keep the old one and the new one can be expressed just
:: by G.AdjacentSet({v}) = {}. Should this be revised?
:: Ask Lorna and Ryan. For loopless graphs it
:: does not matter, see below.
theorem :: CHORD:56 :: AdjacentV01:
for G being loopless _Graph, u being Vertex of G holds G.AdjacentSet({
u}) = {} iff u is isolated;
theorem :: CHORD:57 :: Connected0
for G being _Graph, G0 being Subgraph of G, S being non empty
Subset of the_Vertices_of G, x being Vertex of G, G1 being (inducedSubgraph of
G,S), G2 being (inducedSubgraph of G,S\/{x}) st G1 is connected & x in G
.AdjacentSet(the_Vertices_of G1) holds G2 is connected;
theorem :: CHORD:58 :: Simplicial2a
for G being _Graph for S being non empty Subset of
the_Vertices_of G for H being inducedSubgraph of G,S for u being Vertex of G st
u in S & G.AdjacentSet({u}) c= S for v being Vertex of H st u=v holds G
.AdjacentSet({u}) = H.AdjacentSet({v});
:: Adjacency set as a subgraph of G
definition
let G be _Graph, S be set;
mode AdjGraph of G,S -> Subgraph of G means
:: CHORD:def 5
it is inducedSubgraph of
G,G.AdjacentSet(S) if S is Subset of the_Vertices_of G;
end;
theorem :: CHORD:59 :: AdjGraph00
for G1, G2 be _Graph st G1 == G2 for u1 being Vertex of G1, u2
being Vertex of G2 st u1 = u2 for H1 being AdjGraph of G1,{u1}, H2 being
AdjGraph of G2,{u2} holds H1 == H2;
theorem :: CHORD:60 :: Simplicial2b
for G being _Graph for S being non empty Subset of
the_Vertices_of G for H being inducedSubgraph of G,S for u being Vertex of G st
u in S & G.AdjacentSet({u}) c= S & G.AdjacentSet({u}) <> {} for v being Vertex
of H st u=v for Ga being AdjGraph of G,{u}, Ha being AdjGraph of H,{v} holds Ga
== Ha;
definition
let G be _Graph;
attr G is complete means
:: CHORD:def 6
for u,v being Vertex of G st u <> v holds u, v are_adjacent;
end;
theorem :: CHORD:61 :: Completetr
for G being _Graph st G is trivial holds G is complete;
registration
cluster trivial -> complete for _Graph;
end;
registration
cluster trivial simple complete for _Graph;
cluster non trivial finite simple complete for _Graph;
end;
theorem :: CHORD:62 :: Complete00
for G1,G2 being _Graph st G1 == G2 holds G1 is complete implies
G2 is complete;
theorem :: CHORD:63 :: Complete01
for G being complete _Graph, S being Subset of the_Vertices_of G
for H being inducedSubgraph of G,S holds H is complete;
begin :: Simplicial vertex :: Golumbic p. 81
definition
let G be _Graph, v be Vertex of G;
attr v is simplicial means
:: CHORD:def 7
G.AdjacentSet({v}) <> {} implies for G2
being AdjGraph of G,{v} holds G2 is complete;
end;
theorem :: CHORD:64 :: Simplicial0
for G being complete _Graph, v being Vertex of G holds v is simplicial;
theorem :: CHORD:65 :: Simplicial01
for G being trivial _Graph, v being Vertex of G holds v is simplicial;
theorem :: CHORD:66 :: Simplicial1
for G1,G2 being _Graph st G1 == G2 for u1 being Vertex of G1, u2
being Vertex of G2 st u1=u2 & u1 is simplicial holds u2 is simplicial;
theorem :: CHORD:67 :: Simplicial2
for G being _Graph for S being non empty Subset of
the_Vertices_of G for H being inducedSubgraph of G,S for u being Vertex of G st
u in S & G.AdjacentSet({u}) c= S for v being Vertex of H st u=v holds u is
simplicial iff v is simplicial;
theorem :: CHORD:68 :: Simplicial03
for G being _Graph, v being Vertex of G st v is simplicial
for a,b being object
st a<>b & a in G.AdjacentSet({v}) & b in G.AdjacentSet({v}) holds
ex e being object st e Joins a,b,G;
theorem :: CHORD:69 :: Simplicial03a
for G being _Graph, v being Vertex of G st not v is simplicial
ex a,b being Vertex of G st a<>b & v<>a & v<>b & v,a are_adjacent & v,b
are_adjacent & not a,b are_adjacent;
begin :: Vertex separator :: Golumbic, p. 84
definition
let G be _Graph, a,b be Vertex of G;
assume that
a<>b and
not a,b are_adjacent;
mode VertexSeparator of a,b -> Subset of the_Vertices_of G means
:: CHORD:def 8
not
a in it & not b in it & for G2 being removeVertices of G,it holds not (ex W
being Walk of G2 st W is_Walk_from a,b);
end;
theorem :: CHORD:70 :: VS01
for G being _Graph, a,b being Vertex of G st a<>b & not a,b
are_adjacent for S being VertexSeparator of a,b holds S is VertexSeparator of b
,a;
:: alternate characterization of Vertex Separator
theorem :: CHORD:71 :: VS02
for G being _Graph, a,b being Vertex of G st a<>b & not a,b
are_adjacent for S being Subset of the_Vertices_of G holds S is VertexSeparator
of a,b iff (not a in S & not b in S & for W being Walk of G st W is_Walk_from a
,b holds ex x being Vertex of G st x in S & x in W.vertices());
theorem :: CHORD:72 :: VS07
for G being _Graph, a,b being Vertex of G st a<>b & not a,b
are_adjacent for S being VertexSeparator of a,b for W being Walk of G st W
is_Walk_from a,b ex k being odd Nat st 1 < k & k < len W & W.k in S;
theorem :: CHORD:73 :: VS08a
for G being _Graph, a,b being Vertex of G st a<>b & not a,b
are_adjacent for S being VertexSeparator of a,b st S = {} holds not ex W being
Walk of G st W is_Walk_from a,b;
theorem :: CHORD:74
for G being _Graph, a,b being Vertex of G st a<>b ¬ a,b
are_adjacent & not ex W being Walk of G st W is_Walk_from a,b holds {} is
VertexSeparator of a,b;
theorem :: CHORD:75 :: VS11
for G being _Graph, a,b being Vertex of G st a<>b & not a,b
are_adjacent for S being VertexSeparator of a,b, G2 being removeVertices of G,S
for a2 being Vertex of G2 holds G2.reachableFrom(a2) /\ S = {};
theorem :: CHORD:76
for G being _Graph, a,b being Vertex of G st a<>b & not a,b
are_adjacent for S being VertexSeparator of a,b, G2 being removeVertices of G,S
for a2,b2 being Vertex of G2 st a2=a & b2=b holds G2.reachableFrom(a2) /\ G2
.reachableFrom(b2) = {};
theorem :: CHORD:77 :: VS10
for G being _Graph, a,b being Vertex of G st a<>b & not a,b
are_adjacent for S being VertexSeparator of a,b for G2 being removeVertices of
G,S holds a is Vertex of G2 & b is Vertex of G2;
definition
let G be _Graph, a,b be Vertex of G;
let S be VertexSeparator of a,b;
attr S is minimal means
:: CHORD:def 9
for T being Subset of S st T <> S holds not T is VertexSeparator of a,b;
end;
theorem :: CHORD:78 :: VS000
for G being _Graph, a,b being Vertex of G for S being VertexSeparator
of a,b st S = {} holds S is minimal;
theorem :: CHORD:79 :: minVSexistance
for G being finite _Graph for a,b being Vertex of G ex S being
VertexSeparator of a,b st S is minimal;
theorem :: CHORD:80 :: VS13
:: Property "symmetry" for 2 argument modes could be used if we had it
:: as VertexSeparator of a,b is a VertexSeparator of b,a
:: then this theorem would not be needed
for G being _Graph, a,b being Vertex of G st a<>b & not a,b
are_adjacent for S being VertexSeparator of a,b st S is minimal for T being
VertexSeparator of b,a st S=T holds T is minimal;
theorem :: CHORD:81 :: VS06
for G being _Graph, a,b being Vertex of G st a<>b & not a,b
are_adjacent for S being VertexSeparator of a,b st S is minimal for x being
Vertex of G st x in S ex W being Walk of G st W is_Walk_from a,b & x in W
.vertices();
theorem :: CHORD:82 :: VertexSep0
for G being _Graph for a,b being Vertex of G st a<>b & not a,b
are_adjacent for S being VertexSeparator of a,b st S is minimal for H being
removeVertices of G,S for aa being Vertex of H st aa=a for x being Vertex of G
st x in S ex y being Vertex of G st y in H.reachableFrom(aa) & x,y are_adjacent
;
theorem :: CHORD:83 :: VertexSep01
:: Property "symmetry" for 2 argument modes could be used if we had it
:: as VertexSeparator of a,b is a VertexSeparator of b,a
for G being _Graph for a,b being Vertex of G st a<>b & not a,b
are_adjacent for S being VertexSeparator of a,b st S is minimal for H being
removeVertices of G,S for aa being Vertex of H st aa=b for x being Vertex of G
st x in S ex y being Vertex of G st y in H.reachableFrom(aa) & x,y are_adjacent
;
begin :: Chordal graphs :: Golumbic, p. 81
:: The notion of a chord. Is it worthwhile having it?
:: definition let G be _Graph, W be Walk of G, e be set;
:: pred e is_chord_of W means
:: ex m, n being odd Nat st m < n & n <= len W & W.m <> W.n &
:: e Joins W.m,W.n,G &
:: for f being set st f in W.edges() holds not f Joins W.m,W.n,G;
:: end;
:: More general notion of a chordal Walk. Is such a notion useful? Or
:: should we stick with chordal Path?
definition
let G be _Graph, W be Walk of G;
attr W is chordal means
:: CHORD:def 10
ex m, n being odd Nat st m+2 < n
& n <= len W & W.m <> W.n & (ex e being object st e Joins W.m,W.n,G) &
for f being object st f in W.edges() holds not f Joins W.m,W.n,G;
end;
notation
let G be _Graph, W be Walk of G;
antonym W is chordless for W is chordal;
end;
:: The other characterization of chordal is 'more' technical and
:: sometimes more convenient to work with. Is this really true?
:: I have tried to save as much as possible of what Broderic has already done.
:: Need separate theorems for walks and paths! They cannot be put as an iff.
theorem :: CHORD:84 :: ChordalWalk01
for G being _Graph, W being Walk of G st W is chordal ex m,n
being odd Nat st m+2 < n & n <= len W & W.m <> W.n &
(ex e being object st e Joins W.m,W.n,G) &
(W is Cycle-like implies not (m=1 & n = len W) & not (m
=1 & n = len W-2) & not (m=3 & n = len W));
theorem :: CHORD:85 :: ChordalPath01
for G being _Graph, P being Path of G st ex m,n being odd Nat
st m+2 < n & n <= len P & (ex e being object st e Joins P.m,P.n,G)
& (P is Cycle-like implies not (m=1 & n = len P) & not (m=1 & n = len P-2) &
not (m=3 & n = len P)) holds P is chordal;
theorem :: CHORD:86 :: ChordalWalk02
for G1,G2 being _Graph st G1 == G2 for W1 being Walk of G1, W2
being Walk of G2 st W1=W2 holds W1 is chordal implies W2 is chordal;
theorem :: CHORD:87 :: ChordalWalk03
for G being _Graph, S being non empty Subset of the_Vertices_of
G, H being (inducedSubgraph of G,S), W1 being Walk of G, W2 being Walk of H st
W1 = W2 holds W2 is chordal iff W1 is chordal;
theorem :: CHORD:88
for G being _Graph, W being Walk of G st W is Cycle-like & W is
chordal & W.length()=4
ex e being object st e Joins W.1,W.5,G or e Joins W.3,W.7,G;
theorem :: CHORD:89 :: MinChordal01
for G being _Graph, W being Walk of G st W is minlength holds W is chordless;
theorem :: CHORD:90
for G being _Graph, W being Walk of G st len W = 5 & not W.first(),W
.last() are_adjacent holds W is chordless;
theorem :: CHORD:91
for G being _Graph, W being Walk of G holds W is chordal iff W
.reverse() is chordal;
theorem :: CHORD:92 :: CPath03
for G being _Graph, P being Path of G st P is open & P is
chordless for m,n being odd Nat st m < n & n <= len P holds
(ex e being object st e Joins P.m,P.n,G) iff m+2 = n;
theorem :: CHORD:93
for G being _Graph, P being Path of G st P is open & P is chordless
for m,n being odd Nat st m < n & n <= len P holds P.cut(m,n) is
chordless & P.cut(m,n) is open;
theorem :: CHORD:94
for G being _Graph, S being non empty Subset of the_Vertices_of G, H
being (inducedSubgraph of G,S), W being Walk of G, V being Walk of H st W = V
holds W is chordless iff V is chordless;
definition
let G be _Graph;
attr G is chordal means
:: CHORD:def 11
for P being Walk of G st P.length() > 3 & P is Cycle-like holds P is chordal;
end;
theorem :: CHORD:95 :: Chordal01
for G1,G2 being _Graph st G1 == G2 holds G1 is chordal implies G2 is chordal;
theorem :: CHORD:96 :: Chordal02
for G being finite _Graph st card the_Vertices_of G <= 3 holds G is chordal;
registration
cluster trivial finite chordal for _Graph;
cluster non trivial finite simple chordal for _Graph;
cluster complete -> chordal for _Graph;
end;
registration
let G be chordal _Graph, V be set;
cluster -> chordal for inducedSubgraph of G,V;
end;
theorem :: CHORD:97
for G being chordal _Graph, P being Path of G st P is open & P is
chordless
for x,e being object st (not x in P.vertices() & e Joins P.last(),x,G &
not ex f being object st f Joins P.(len P-2),x,G)
holds P.addEdge(e) is Path-like
& P.addEdge(e) is open & P.addEdge(e) is chordless;
:: Golumbic, page 83. Theorem 4.1 (i) ==> (iii)
theorem :: CHORD:98
for G being chordal _Graph, a,b being Vertex of G st a<>b & not
a,b are_adjacent for S being VertexSeparator of a,b st S is minimal & S is non
empty for H being inducedSubgraph of G,S holds H is complete;
:: Golumbic, page 83, Theorem 4.1 (iii)->(i)
theorem :: CHORD:99
for G being finite _Graph st for a,b being Vertex of G st a<>b & not a
,b are_adjacent for S being VertexSeparator of a,b st S is minimal & S is non
empty for G2 being inducedSubgraph of G,S holds G2 is complete holds G is
chordal;
:: Exercise 12, p. 101.
:: This needs "finite-branching", we do it for finite though
theorem :: CHORD:100
for G being finite chordal _Graph, a, b being Vertex of G st a
<> b & not a,b are_adjacent for S being VertexSeparator of a,b st S is minimal
for H being removeVertices of G,S, a1 being Vertex of H st a = a1 ex c being
Vertex of G st c in H.reachableFrom(a1) & for x being Vertex of G st x in S
holds c,x are_adjacent;
theorem :: CHORD:101
for G being finite chordal _Graph, a, b being Vertex of G st a <> b &
not a,b are_adjacent for S being VertexSeparator of a,b st S is minimal for H
being removeVertices of G,S, a1 being Vertex of H st a = a1 for x, y being
Vertex of G st x in S & y in S holds ex c being Vertex of G st c in H
.reachableFrom(a1) & c,x are_adjacent & c,y are_adjacent;
:: Golumbic, page 83, Lemma 4.2.
theorem :: CHORD:102
for G being non trivial finite chordal _Graph st not G is
complete ex a,b being Vertex of G st a<>b & not a,b are_adjacent & a is
simplicial & b is simplicial;
theorem :: CHORD:103
for G being finite chordal _Graph ex v being Vertex of G st v is simplicial;
begin :: Vertex Elimination Scheme :: Golumbic, p. 82
definition
let G be finite _Graph;
mode VertexScheme of G -> FinSequence of the_Vertices_of G means
:: CHORD:def 12
it is one-to-one & rng it = the_Vertices_of G;
end;
registration
let G be finite _Graph;
cluster -> non empty for VertexScheme of G;
end;
theorem :: CHORD:104
for G being finite _Graph, S being VertexScheme of G holds len S =
card the_Vertices_of G;
theorem :: CHORD:105
for G being finite _Graph, S being VertexScheme of G holds 1 <= len S;
theorem :: CHORD:106
for G, H being finite _Graph, g being VertexScheme of G st G ==
H holds g is VertexScheme of H;
definition
let G be finite _Graph, S be VertexScheme of G, x be Vertex of G;
redefine func x..S -> non zero Element of NAT;
end;
definition
let G be finite _Graph, S be VertexScheme of G, n be Nat;
redefine func S.followSet(n) -> Subset of the_Vertices_of G;
end;
theorem :: CHORD:107 :: NeVSchFol:
for G being finite _Graph, S being VertexScheme of G, n being
non zero Nat st n <= len S holds S.followSet(n) is non empty;
definition
let G be finite _Graph, S be VertexScheme of G;
attr S is perfect means
:: CHORD:def 13
for n being non zero Nat st n <=
len S for Gf being inducedSubgraph of G,S.followSet(n) for v being Vertex of Gf
st v = S.n holds v is simplicial;
end;
:: finite is needed unless we add loopless
theorem :: CHORD:108
for G being finite trivial _Graph, v being Vertex of G ex S
being VertexScheme of G st S = <*v*> & S is perfect;
theorem :: CHORD:109
for G being finite _Graph, V being VertexScheme of G holds V is
perfect iff for a,b,c being Vertex of G st b<>c & a,b are_adjacent & a,c
are_adjacent for va,vb,vc being Nat st va in dom V & vb in dom V &
vc in dom V & V.va = a & V.vb = b & V.vc = c & va < vb & va < vc holds b,c
are_adjacent;
:: Golubmic pg 83-84, Theorem 4.1 (i) ==> (ii)
registration
let G be finite chordal _Graph;
cluster perfect for VertexScheme of G;
end;
theorem :: CHORD:110
for G, H being finite chordal _Graph, g being perfect VertexScheme of
G st G == H holds g is perfect VertexScheme of H;
:: Chordal41c:
:: Golubmic pg 83-84, Theorem 4.1 (ii) ==> (i)
theorem :: CHORD:111
for G being finite _Graph st ex S being VertexScheme of G st S is
perfect holds G is chordal;