Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
C
Comparable box dimension
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Deploy
Releases
Package registry
Model registry
Operate
Terraform modules
Monitor
Incidents
Analyze
Value stream analytics
Contributor analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
GitLab community forum
Contribute to GitLab
Provide feedback
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
Zdenek Dvorak
Comparable box dimension
Commits
5ada1937
Commit
5ada1937
authored
3 years ago
by
Zdenek Dvorak
Browse files
Options
Downloads
Patches
Plain Diff
Added details to the proof of Lemma 10.
parent
0d75f91f
No related branches found
No related tags found
No related merge requests found
Changes
1
Show whitespace changes
Inline
Side-by-side
Showing
1 changed file
comparable-box-dimension.tex
+256
-198
256 additions, 198 deletions
comparable-box-dimension.tex
with
256 additions
and
198 deletions
comparable-box-dimension.tex
+
256
−
198
View file @
5ada1937
...
...
@@ -210,58 +210,65 @@ Then $h$ is a touching representation of $G$ by comparable boxes in $\mathbb{R}^
Let
$
G
\boxtimes
H
$
denote the
\emph
{
strong product
}
of the graphs
$
G
$
and
$
H
$
, i.e., the graph with vertex set
$
V
(
G
)
\times
V
(
H
)
$
and with
distinct vertices
$
(
u
_
1
,v
_
1
)
$
and
$
(
u
_
2
,v
_
2
)
$
adjacent if and only if,
either
$
u
_
1
=
u
_
2
$
or
$
u
_
1
u
_
2
\in
E
(
G
)
$
, and either
$
v
_
1
=
v
_
2
$
or
$
v
_
1
v
_
2
\in
E
(
G
)
$
. To obtain a touching representation of
$
G
\boxtimes
H
$
it suffice to take a product of representations of
$
G
$
and
$
H
$
, but
the obtained representation may contain uncomparable boxes. Thus,
bounding
$
\cbdim
(
G
\boxtimes
H
)
$
in terms of
$
\cbdim
(
G
)
$
and
$
\cbdim
(
H
)
$
seems to be a complicated task. In the following lemma we
overcome this issue, by constraining one of the representations.
distinct vertices
$
(
u
_
1
,v
_
1
)
$
and
$
(
u
_
2
,v
_
2
)
$
adjacent if and only if
$
u
_
1
$
is equal to or adjacent to
$
u
_
2
$
in
$
G
$
and
$
v
_
1
$
is equal to or adjacent to
$
v
_
2
$
in
$
H
$
.
To obtain a touching representation of
$
G
\boxtimes
H
$
it suffices to take a product of representations of
$
G
$
and
$
H
$
, but
the resulting representation may contain incomparable boxes.
Indeed,
$
\cbdim
(
G
\boxtimes
H
)
$
in general is not bounded by a function
of
$
\cbdim
(
G
)
$
and
$
\cbdim
(
H
)
$
; for example, every star has comparable box dimension
at most two, but the strong product of the star
$
K
_{
1
,n
}$
with itself contains
$
K
_{
n,n
}$
as an induced subgraph, and thus its comparable box dimension is at least
$
\Omega
(
\log
n
)
$
.
However, as shown in the following lemma, this issue does not arise if the representation of
$
H
$
consists of translates
of a single box; by scaling, we can without loss of generality assume this box is a unit hypercube.
\begin{lemma}
\label
{
lemma-sp
}
Consider a graph
$
H
$
having a touching representation
$
h
$
in
$
\mathbb
{
R
}^{
d
_
H
}$
with
hypercubes of unit size. Then for any graph
$
G
$
, the strong product of these graphs
is such tha
t
$
\cbdim
(
G
\boxtimes
H
)
\le
\cbdim
(
G
)
+
d
_
H
$
.
$
\mathbb
{
R
}^{
d
_
H
}$
by axis-aligned
hypercubes of unit size. Then for any graph
$
G
$
, the strong product
$
G
\boxtimes
H
$
of these graphs
has comparable box dimension at mos
t
$
\cbdim
(
G
)
+
d
_
H
$
.
\end{lemma}
\begin{proof}
The proof simply consists in taking a product of the two
representations. Indeed, consider a touching respresentation
with
comparable boxes
$
g
$
of
$
G
$
in
$
\mathbb
{
R
}^{
d
_
G
}$
, with
representations. Indeed, consider a touching respresentation
$
g
$
of
$
G
$
by
comparable boxes in
$
\mathbb
{
R
}^{
d
_
G
}$
, with
$
d
_
G
=
\cbdim
(
G
)
$
, and the representation
$
h
$
of
$
H
$
. Let us define a
representation
$
f
$
of
$
G
\boxtimes
H
$
in
$
\mathbb
{
R
}^{
d
_
G
+
d
_
H
}$
as
follows.
$$
f
((
u,v
))[
i
]=
\begin
{
cases
}
g
(
u
)[
i
]
&
\text
{
if $i
\le
d
_
G$
}
\\
h
(
u
)[
i
-
d
_
G
]
&
\text
{
if $i > d
_
G$
}
h
(
v
)[
i
-
d
_
G
]
&
\text
{
if $i > d
_
G$
}
\end
{
cases
}$$
Notice first that the boxes of
$
f
$
are comparable as
$
f
((
u,v
))
\sqsubset
f
((
u',v'
))
$
if and only if
$
g
(
u
)
\sqsubset
g
(
u'
)
$
.
Now let us observe that for any two vertices
$
u, u'
$
of
$
G
$
, there
is an hyperplane separating the interiors of
$
g
(
u
)
$
and
$
g
(
u'
)
$
, and
similarly for
$
h
$
and
$
H
$
. This implies that the boxes in
$
f
$
are
interiorly disjoint. Indeed, the same hyperplane that separates
$
g
(
u
)
$
and
$
g
(
u'
)
$
, when extended to
$
\mathbb
{
R
}^{
d
_
G
+
d
_
H
}$
, now separates
any two boxes
$
f
((
u,v
))
$
and
$
f
((
u',v'
))
$
. This implies that
$
f
$
is
a touching representation of a subgraph of
$
G
\boxtimes
H
$
.
Similarly, one can also observe that there is a point
$
p
$
in the
intersection of
$
f
((
u,v
))
$
and
$
f
((
u',v'
))
$
, if and only if there is
a point
$
p
_
G
$
in the intersection of
$
g
(
u
)
$
and
$
g
(
u'
)
$
, and a point
$
p
_
H
$
in the intersection of
$
h
(
v
)
$
and
$
h
(
v'
)
$
. Indeed, one can
obtain
$
p
_
G
$
and
$
p
_
H
$
by projecting
$
p
$
in
$
\mathbb
{
R
}^{
d
_
G
}$
or in
$
\mathbb
{
R
}^{
d
_
H
}$
respectively, and conversely
$
p
$
can be obtained
by taking the product of
$
p
_
G
$
and
$
p
_
H
$
. Thus
$
f
$
is indeed a
touching representation of
$
G
\boxtimes
H
$
.
Consider distinct vertices
$
(
u,v
)
$
and
$
(
u',v'
)
$
of
$
G
\boxtimes
H
$
.
The boxes
$
g
(
u
)
$
and
$
g
(
u'
)
$
are comparable, say
$
g
(
u
)
\sqsubseteq
g
(
u'
)
$
. Since
$
h
(
v'
)
$
is a translation of
$
h
(
v
)
$
, this implies that
$
f
((
u,v
))
\sqsubseteq
f
((
u',v'
))
$
. Hence, the boxes
of the representation
$
f
$
are pairwise comparable.
The boxes of the representations
$
g
$
and
$
h
$
have pairwise disjoint interiors.
Hence, if
$
u
\neq
u'
$
, then there exists
$
i
\le
d
_
G
$
such that the interiors
of the intervals
$
f
((
u,v
))[
i
]=
g
(
u
)[
i
]
$
and
$
f
((
u',v'
))[
i
]=
g
(
u'
)[
i
]
$
are disjoint;
and if
$
v
\neq
v'
$
, then there exists
$
i
\le
d
_
H
$
such that the interiors
of the intervals
$
f
((
u,v
))[
i
+
d
_
G
]=
h
(
v
)[
i
]
$
and
$
f
((
u',v'
))[
i
+
d
_
G
]=
h
(
v'
)[
i
]
$
are disjoint.
Consequently, the interiors of boxes
$
f
((
u,v
))
$
and
$
f
((
u',v'
))
$
are pairwise disjoint.
Moreover, if
$
u
\neq
u'
$
and
$
uu'
\not\in
E
(
G
)
$
, or if
$
v
\neq
v'
$
and
$
vv'
\not\in
E
(
G
)
$
,
then the intervals discussed above (not just their interiors) are disjoint for some
$
i
$
;
hence, if
$
(
u,v
)
$
and
$
(
u',v'
)
$
are not adjacent in
$
G
\boxtimes
H
$
, then
$
f
((
u,v
))
\cap
f
((
u',v'
))=
\emptyset
$
.
Therefore,
$
f
$
is a touching representation of a subgraph of
$
G
\boxtimes
H
$
.
Finally, suppose that
$
(
u,v
)
$
and
$
(
u',v'
)
$
are adjacent in
$
G
\boxtimes
H
$
.
Then there exists a point
$
p
_
G
$
in the intersection of
$
g
(
u
)
$
and
$
g
(
u'
)
$
,
since
$
u
=
u'
$
or
$
uu'
\in
E
(
G
)
$
and
$
g
$
is a touching representation of
$
G
$
;
and similarly, there exists a point
$
p
_
H
$
in the intersection of
$
h
(
v
)
$
and
$
h
(
v'
)
$
.
Then
$
p
_
G
\times
p
_
H
$
is a point in the intersection of
$
f
((
u,v
))
$
and
$
f
((
u',v'
))
$
.
Hence,
$
f
$
is indeed a touching representation of
$
G
\boxtimes
H
$
.
\end{proof}
\subsection
{
S
ubgraph
}
\subsection
{
Taking a s
ubgraph
}
Examples show that t
he comparable box dimension of a graph
$
G
$
may be
larger than the one a subgraph
$
H
$
of
$
G
$
. However we show that the
T
he comparable box dimension of a
subgraph of a
graph
$
G
$
may be
larger than
$
\cbdim
(
G
)
$
, see the end of this
section for an example
. However
,
we show that the
comparable box dimension of a subgraph is at most exponential in the
comparable box dimension of the whole graph. This is essentially
Corollary~25 in~
\cite
{
subconvex
}
, but since the setting is somewhat
...
...
@@ -273,24 +280,31 @@ If $G$ is a subgraph of a graph $G'$, then $\cbdim(G)\le \cbdim(G')+\chi^2_s(G')
\end{lemma}
\begin{proof}
As we can remove the boxes that represent the vertices, we can assume
$
V
(
G'
)=
V
(
G
)
$
.
Let
$
f
$
be a touching representation by comparable boxes in
$
\mathbb
{
R
}^
d
$
, where
$
d
=
\cbdim
(
G'
)
$
. Let
$
\varphi
$
Let
$
f
$
be a touching representation
of
$
G'
$
by comparable boxes in
$
\mathbb
{
R
}^
d
$
, where
$
d
=
\cbdim
(
G'
)
$
. Let
$
\varphi
$
be a star coloring of
$
G'
$
using colors
$
\{
1
,
\ldots
,c
\}
$
, where
$
c
=
\chi
_
s
(
G'
)
$
.
For any distinct colors
$
i,j
\in\{
1
,
\ldots
,c
\}
$
, let
$
A
_{
i,j
}
\subseteq
V
(
G
)
$
consist of vertices
$
u
$
of color
$
i
$
such that there exists a vertex
$
v
$
of color
$
j
$
such that
$
uv
\in
E
(
G'
)
$
and
$
uv
\not\in
E
(
G
)
$
.
For any distinct colors
$
i,j
\in\{
1
,
\ldots
,c
\}
$
, let
$
A
_{
i,j
}
\subseteq
V
(
G
)
$
be the set of vertices
$
u
$
of color
$
i
$
such that there exists a vertex
$
v
$
of color
$
j
$
such that
$
uv
\in
E
(
G'
)
\setminus
E
(
G
)
$
. For each
$
u
\in
A
_{
i,j
}$
,
let
$
a
_
j
(
u
)
$
denote such a vertex
$
v
$
chosen arbitrarily.
Let us define a representation
$
h
$
by boxes in
$
\mathbb
{
R
}^{
d
+
\binom
{
c
}{
2
}}$
by starting from the representation
$
f
$
and,
for each pair
$
i<j
$
of colors, adding a dimension
$
d
_{
i,j
}$
and setting
$
h
(
v
)[
d
_{
i,j
}
]=[
1
/
3
,
4
/
3
]
$
for
$
v
\in
A
_{
i,j
}$
,
$
h
(
v
)[
d
_{
i,j
}
]=[-
4
/
3
,
-
1
/
3
]
$
for
$
v
\in
A
_{
j,i
}$
,
and
$
h
(
v
)[
d
_{
i,j
}
](
v
)=[-
1
/
2
,
1
/
2
]
$
otherwise. Note that the boxes in this extended representation are comparable,
$$
h
(
v
)[
d
_{
i,j
}
]=
\begin
{
cases
}
[
1
/
3
,
4
/
3
]
&
\text
{
if $v
\in
A
_{
i,j
}
$
}
\\
[-
4
/
3
,
-
1
/
3
]
&
\text
{
if $v
\in
A
_{
j,i
}
$
}
\\
[-
1
/
2
,
1
/
2
]
&
\text
{
otherwise.
}
\end
{
cases
}$$
Note that the boxes in this extended representation are comparable,
as in the added dimensions, all the boxes have size
$
1
$
.
Suppose
$
uv
\in
E
(
G
)
$
, where
$
\varphi
(
u
)=
i
$
and
$
\varphi
(
v
)=
j
$
and say
$
i<j
$
. The boxes
$
f
(
u
)
$
and
$
f
(
v
)
$
touch.
We cannot have
$
u
\in
A
_{
i,j
}$
and
$
v
\in
A
_{
j,u
}$
, as then
$
G'
$
would contain a 4-vertex path in colors
$
i
$
and
$
j
$
.
Hence, in any added dimension
$
d'
$
, at least one of
$
h
(
u
)
$
and
$
h
(
v
)
$
is represented by the interval
$
[-
1
/
2
,
1
/
2
]
$
,
and thus
$
h
(
u
)[
d'
]
\cap
h
(
v
)[
d'
]
\neq\emptyset
$
. Therefore, the boxes
$
h
(
u
)
$
and
$
h
(
v
)
$
touch.
Suppose
$
uv
\in
E
(
G
)
$
, where
$
\varphi
(
u
)=
i
$
and
$
\varphi
(
v
)=
j
$
and say
$
i<j
$
.
We cannot have
$
u
\in
A
_{
i,j
}$
and
$
v
\in
A
_{
j,i
}$
, as then
$
a
_
j
(
u
)
uva
_
i
(
v
)
$
would be a 4-vertex path in
$
G'
$
in colors
$
i
$
and
$
j
$
.
Hence, in any added dimension
$
d'
$
, we have
$
h
(
u
)[
d'
]=[-
1
/
2
,
1
/
2
]
$
or
$
h
(
v
)[
d'
]=[-
1
/
2
,
1
/
2
]
$
,
and thus
$
h
(
u
)[
d'
]
\cap
h
(
v
)[
d'
]
\neq\emptyset
$
.
Since the boxes
$
f
(
u
)
$
and
$
f
(
v
)
$
touch, it follows that the boxes
$
h
(
u
)
$
and
$
h
(
v
)
$
touch as well.
Suppose now that
$
uv
\not\in
E
(
G
)
$
. If
$
uv
\not\in
E
(
G'
)
$
, then
$
f
(
u
)
$
is disjoint from
$
f
(
v
)
$
, and thus
$
h
(
u
)
$
is disjoint from
$
h
(
v
)
$
. Hence, we can assume
$
uv
\in
E
(
G'
)
$
,
$
\varphi
(
u
)=
i
$
,
$
\varphi
(
v
)=
j
$
and
$
i<j
$
. Then
$
u
\in
A
_{
i,j
}$
,
$
v
\in
A
_{
j,i
}$
,
$
h
(
v
)
$
. Hence, we can assume
$
uv
\in
E
(
G'
)
\setminus
E
(
G
)
$
,
$
\varphi
(
u
)=
i
$
,
$
\varphi
(
v
)=
j
$
and
$
i<j
$
. Then
$
u
\in
A
_{
i,j
}$
,
$
v
\in
A
_{
j,i
}$
,
$
h
(
u
)[
d
_{
i,j
}
]=[
1
/
3
,
4
/
3
]
$
,
$
h
(
v
)[
d
_{
j,i
}
]=[-
4
/
3
,
-
1
/
3
]
$
, and
$
h
(
u
)
\cap
h
(
v
)=
\emptyset
$
.
Consequently,
$
h
$
is a touching representation of
$
G
$
by comparable boxes in dimension
$
d
+
\binom
{
c
}{
2
}
\le
d
+
c
^
2
$
.
...
...
@@ -302,7 +316,7 @@ Let us now combine Lemmas~\ref{lemma-chrom} and \ref{lemma-subg}.
If
$
G
$
is a subgraph of a graph
$
G'
$
, then
$
\cbdim
(
G
)
\le
\cbdim
(
G'
)+
4
\cdot
81
^{
\cbdim
(
G'
)
}
\le
5
\cdot
81
^{
\cbdim
(
G'
)
}$
.
\end{corollary}
Let us remark that an exponential increase in the dimension is unavoidable: We have
$
\cbdim
{
K
_{
2
^
d
}
}
=
d
$
,
Let us remark that an exponential increase in the dimension is unavoidable: We have
$
\cbdim
(
K
_{
2
^
d
}
)
=
d
$
,
but the graph obtained from
$
K
_{
2
^
d
}$
by deleting a perfect matching has comparable box dimension
$
2
^{
d
-
1
}$
.
...
...
@@ -324,162 +338,206 @@ an arbitrary number of clique-sums. We thus introduce the notion of
\emph
{
clique-sum extendable
}
representations.
\begin{definition}
Consider a graph
$
G
$
with a distinguished clique
$
C
^
*
$
, called the
\emph
{
root clique
}
of
$
G
$
. A touching representation
(with comparable
b
oxes or not)
$
h
$
of
$
G
$
in
$
\mathbb
{
R
}^
d
$
is called
\emph
{$
C
^
*
$
-clique-sum extendable
}
if the following conditions hold.
Consider a graph
$
G
$
with a distinguished clique
$
C
^
\star
$
, called the
\emph
{
root clique
}
of
$
G
$
. A touching representation
$
h
$
of
$
G
$
b
y (not necessarily comparable) boxes
in
$
\mathbb
{
R
}^
d
$
is called
\emph
{$
C
^
\star
$
-clique-sum extendable
}
if the following conditions hold
for every sufficiently small
$
\varepsilon
>
0
$
.
\begin{itemize}
\item
[{\bf(vertices)}]
There are
$
|V
(
C
^
*)
|
$
dimensions, denoted
$
d
_
u
$
for each
vertex
$
u
\in
V
(
C
^
*)
$
,
such that:
\item
[{\bf(vertices)}]
For each
$
u
\in
V
(
C
^
\star
)
$
, there exists a dimension
$
d
_
u
$
,
such that:
\begin{itemize}
\item
[(v1)]
for each vertex
$
u
\in
V
(
C
^
*)
$
,
$
h
(
u
)[
d
_
u
]
=
[-
1
,
0
]
$
and
$
h
(
u
)[
i
]
=
[
0
,
1
]
$
, for any dimension
$
i
\neq
d
_
u
$
, and
\item
[(v2)]
for any vertex
$
v
\notin
V
(
C
^
*)
$
,
$
h
(
v
)
\subset
[
0
,
1
)
^
d
$
.
\item
[(v0)]
$
d
_
u
\neq
d
_{
u'
}$
for distinct
$
u,u'
\in
V
(
C
^
\star
)
$
,
\item
[(v1)]
each vertex
$
u
\in
V
(
C
^
\star
)
$
satisfies
$
h
(
u
)[
d
_
u
]
=
[-
1
,
0
]
$
and
$
h
(
u
)[
i
]
=
[
0
,
1
]
$
for any dimension
$
i
\neq
d
_
u
$
, and
\item
[(v2)]
each vertex
$
v
\notin
V
(
C
^
\star
)
$
satisfies
$
h
(
v
)
\subset
[
0
,
1
)
^
d
$
.
\end{itemize}
\item
[{\bf(cliques)}]
For every clique
$
C
$
of
$
G
$
we define a point
$
p
(
C
)
\in
I
_
C
\cap
[
0
,
1
)
^
d
$
, where
$
I
_
C
=
\
cap
_{
v
\in
V
(
C
)
}
h
(
v
)
$
, and
we define th
e box
$
h
^
\epsilon
(
C
)
$
, for any
$
\epsilon
>
0
$
, by
$
h
^
\epsilon
(
C
)[
i
]
=
[
p
(
C
)[
i
]
,p
(
C
)[
i
]+
\epsilon
]
$
, for every dimension
$
i
$
. Furthermore, for a sufficiently small
$
\epsilon
>
0
$
these
\emph
{
clique boxes
}
verify
the following conditions
.
\item
[{\bf(cliques)}]
For every clique
$
C
$
of
$
G
$
, there exists
a point
$$
p
(
C
)
\in
[
0
,
1
)
^
d
\cap
\big
cap
_{
v
\in
V
(
C
)
}
h
(
v
)
$
$
such that, defining the
\emph
{
cliqu
e box
}
$
h
^
\
var
epsilon
(
C
)
$
by setting
$
$
h
^
\varepsilon
(
C
)[
i
]
=
[
p
(
C
)[
i
]
,p
(
C
)[
i
]+
\varepsilon
]
$$
for every dimension
$
i
$
,
the following conditions
are satisfied:
\begin{itemize}
\item
[(c1)]
For any two cliques
$
C
_
1
\neq
C
_
2
$
,
$
h
^
\epsilon
(
C
_
1
)
\cap
h
^
\epsilon
(
C
_
2
)
=
\emptyset
$
(
i.e.
$
p
(
C
_
1
)
\neq
p
(
C
_
2
)
$
).
\item
[(c2)]
A box
$
h
(
v
)
$
intersects
$
h
^
\epsilon
(
C
)
$
if and only if
\item
[(c1)]
For any two cliques
$
C
_
1
\neq
C
_
2
$
,
$
h
^
\
var
epsilon
(
C
_
1
)
\cap
h
^
\
var
epsilon
(
C
_
2
)
=
\emptyset
$
(
equivalently,
$
p
(
C
_
1
)
\neq
p
(
C
_
2
)
$
).
\item
[(c2)]
A box
$
h
(
v
)
$
intersects
$
h
^
\
var
epsilon
(
C
)
$
if and only if
$
v
\in
V
(
C
)
$
, and in that case their intersection is a facet of
$
h
^
\epsilon
(
C
)
$
incident to
$
p
(
C
)
$
(i.e. if we denote this
intersection
$
I
$
, then
$
I
[
i
]
=
\{
p
(
C
)[
i
]
\}
$
for some dimension
$
i
$
, and
$
I
[
j
]
=
[
p
(
C
)[
j
]
,p
(
C
)[
j
]+
\epsilon
]
$
for the other
dimensions
$
j
\neq
i
$
).
$
h
^
\varepsilon
(
C
)
$
incident to
$
p
(
C
)
$
. That is, there exists a dimension
$
i
_{
C,v
}$
such that for each dimension
$
j
$
,
$$
h
(
v
)[
j
]
\cap
h
^
\varepsilon
(
C
)[
j
]
=
\begin
{
cases
}
\{
p
(
C
)[
i
_{
C,v
}
]
\}
&
\text
{
if $j
=
i
_{
C,v
}
$
}
\\
[
p
(
C
)[
j
]
,p
(
C
)[
j
]+
\varepsilon
]
&
\text
{
otherwise.
}
\end
{
cases
}$$
\end{itemize}
\end{itemize}
\end{definition}
Note that
we may consider that
the root clique
is
empty, that is the
Note that the root clique
can be
empty, that is the
empty subgraph with no vertices. In that case the clique is denoted
$
\emptyset
$
. Let
$
\ecbdim
(
G
)
$
be the minimum dimension such that
$
G
$
has a
$
\emptyset
$
-clique-sum extendable touching representation by
comparable boxes. The following lemma ensures that clique-sum
has an
$
\emptyset
$
-clique-sum extendable touching representation by
comparable boxes.
Let us remark that a clique-sum extendable representation in dimension
$
d
$
implies
such a representation in higher dimensions as well.
\begin{lemma}
\label
{
lemma-add
}
Let
$
G
$
be a graph with a root clique
$
C
^
\star
$
and let
$
h
$
be
a
$
C
^
\star
$
-clique-sum extendable touching representation of
$
G
$
by comparable boxes in
$
\mathbb
{
R
}^
d
$
.
Then
$
G
$
has such a representation in
$
\mathbb
{
R
}^{
d'
}$
for every
$
d'
\ge
d
$
.
\end{lemma}
\begin{proof}
It clearly suffices to consider the case that
$
d'
=
d
+
1
$
.
Note that the
\textbf
{
(vertices)
}
conditions imply that
$
h
(
v'
)
\sqsubseteq
h
(
v
)
$
for every
$
v'
\in
V
(
G
)
\setminus
V
(
C
^
\star
)
$
and
$
v
\in
V
(
C
^
\star
)
$
. We extend the representation
$
h
$
by setting
$
h
(
v
)[
d
+
1
]
=
[
0
,
1
]
$
for
$
v
\in
V
(
C
^
\star
)
$
and
$
h
(
v
)[
d
+
1
]
=
[
0
,
\frac
12
]
$
for
$
v
\in
V
(
G
)
\setminus
V
(
C
^
\star
)
$
.
The clique point
$
p
(
C
)
$
of each clique
$
C
$
is extended by setting
$
p
(
C
)[
d
+
1
]
=
\frac
14
$
.
It is easy to verify that the resulting representation is
$
C
^
\star
$
-clique-sum extendable.
\end{proof}
The following lemma ensures that clique-sum
extendable representations behave well with respect to full
clique-sums.
\begin{lemma}
\label
{
lem-cs
}
Consider two graphs
$
G
_
1
$
and
$
G
_
2
$
, given with a
$
C
^
*
_
1
$
- and a
$
C
^
*
_
2
$
-clique-sum extendable representations
with
comparable boxes
$
h
_
1
$
and
$
h
_
2
$
,
in
$
\mathbb
{
R
}^{
d
_
1
}$
and
$
\mathbb
{
R
}^{
d
_
2
}$
respectively. Let
$
G
$
be the graph obtained
after
performing a full
Consider two graphs
$
G
_
1
$
and
$
G
_
2
$
, given with a
$
C
^
\star
_
1
$
- and a
$
C
^
\star
_
2
$
-clique-sum extendable representations
$
h
_
1
$
and
$
h
_
2
$
by
comparable boxes
in
$
\mathbb
{
R
}^{
d
_
1
}$
and
$
\mathbb
{
R
}^{
d
_
2
}$
,
respectively. Let
$
G
$
be the graph obtained
by
performing a full
clique-sum of these two graphs on any clique
$
C
_
1
$
of
$
G
_
1
$
, and on
the root clique
$
C
^
*
_
2
$
of
$
G
_
2
$
. Then
$
G
$
admits a
$
C
^
*
_
1
$
-clique
sum extendable representation by comparable boxes
$
h
$
in
the root clique
$
C
^
\star
_
2
$
of
$
G
_
2
$
. Then
$
G
$
admits a
$
C
^
\star
_
1
$
-clique
sum extendable representation
$
h
$
by comparable boxes in
$
\mathbb
{
R
}^{
\max
(
d
_
1
,d
_
2
)
}$
.
\end{lemma}
\begin{proof}
By Lemma~
\ref
{
lemma-add
}
, we can assume that
$
d
_
1
=
d
_
2
$
; let
$
d
=
d
_
1
$
.
The idea is to translate (allowing also exchanges of dimensions) and
scale
$
h
_
2
$
to fit in
$
h
_
1
^
\epsilon
(
C
_
1
)
$
. Consider an
$
\epsilon
>
0
$
sufficiently small so that,
$
h
_
1
^
\epsilon
(
C
_
1
)
$
verifies all the
(cliques) conditions, and such that
$
h
_
1
^
\epsilon
(
C
_
1
)
\sqsubseteq
h
_
1
(
v
)
$
for any vertex
$
v
\in
V
(
G
_
1
)
$
. Without loss of generality,
let us assume that
$
V
(
C
_
1
)=
\{
v
_
1
,
\ldots
,v
_
k
\}
$
, and we also assume
that
$
h
_
1
(
v
_
i
)
$
and
$
h
_
1
^
\epsilon
(
C
_
1
)
$
touch in dimension
$
i
$
(i.e.
$
h
_
1
(
v
_
i
)[
i
]
\cap
h
_
1
^
\epsilon
(
C
_
1
)[
i
]
=
\{
p
(
C
_
1
)[
i
]
\}
$
, and
$
h
_
1
(
v
_
i
)[
j
]
\cap
h
_
1
^
\epsilon
(
C
_
1
)[
j
]
=
[
p
(
C
_
1
)[
j
]
,p
(
C
_
1
)[
j
]+
\epsilon
]
$
for
$
j
\neq
i
$
.
scale
$
h
_
2
$
to fit in
$
h
_
1
^
\varepsilon
(
C
_
1
)
$
. Consider an
$
\varepsilon
>
0
$
sufficiently small so that
$
h
_
1
^
\varepsilon
(
C
_
1
)
$
satisfies all the
\textbf
{
(cliques)
}
conditions, and such that
$
h
_
1
^
\varepsilon
(
C
_
1
)
\sqsubseteq
h
_
1
(
v
)
$
for any vertex
$
v
\in
V
(
G
_
1
)
$
. Let
$
V
(
C
_
1
)=
\{
v
_
1
,
\ldots
,v
_
k
\}
$
;
without loss of generality, we can assume
$
i
_{
C
_
1
,v
_
i
}
=
i
$
for
$
i
\in\{
1
,
\ldots
,k
\}
$
,
and thus
$$
h
_
1
(
v
_
i
)[
j
]
\cap
h
_
1
^
\varepsilon
(
C
_
1
)[
j
]
=
\begin
{
cases
}
\{
p
_
1
(
C
_
1
)[
i
]
\}
&
\text
{
if $j
=
i$
}
\\
[
p
_
1
(
C
_
1
)[
j
]
,p
_
1
(
C
_
1
)[
j
]+
\varepsilon
]
&
\text
{
otherwise.
}
\end
{
cases
}$$
Now let us consider
$
G
_
2
$
and its representation
$
h
_
2
$
. Here the
vertices of
$
C
^
*
_
2
$
are also denoted
$
v
_
1
,
\ldots
,v
_
k
$
, and let us
denote
$
d
_{
v
_
i
}$
the dimension in
$
h
_
2
$
that fulfills condition (v1)
with respect to
$
v
_
i
$
.
Let
$
d
=
\max
(
d
_
1
,d
_
2
)
$
. We are now ready for defining
$
h
$
. For the
vertices of
$
G
_
1
$
it is almost the same representation as
$
h
_
1
$
, as
we set
$
h
(
v
)[
i
]
=
h
_
1
(
v
)[
i
]
$
for any dimension
$
i
\le
d
_
1
$
. If
$
d
=
d
_
2
> d
_
1
$
, then for any dimension
$
i>d
_
1
$
we set
$
h
(
v
)[
i
]
=
[
0
,
1
]
$
if
$
v
\in
V
(
C
^
*
_
1
)
$
, and
$
h
(
v
)[
i
]
=
[
0
,
\frac
12
]
$
if
$
v
\in
V
(
G
_
1
)
\setminus
V
(
C
^
*
_
1
)
$
. Similarly the clique points
$
p
_
1
(
C
)
$
become
$
p
(
C
)
$
by setting
$
p
(
C
)[
i
]
=
p
_
1
(
C
)[
i
]
$
for
$
i
\le
d
_
1
$
, and
$
p
(
C
)[
i
]
=
\frac
14
$
for
$
i> d
_
1
$
.
For
$
h
_
2
$
and the vertices in
$
V
(
G
_
2
)
\setminus
\{
v
_
1
,
\ldots
,v
_
k
\}
$
we have to consider a mapping
$
\sigma
$
from
$
\{
1
,
\ldots
,d
_
2
\}
$
to
$
\{
1
,
\ldots
,d
\}
$
such that
$
\sigma
(
d
_{
v
_
i
}
)
=
i
$
. This mapping
describes the changes of dimension we have to perform. We also have
to perform a scaling in order to make
$
h
_
2
$
fit inside
$
h
_
1
^
\epsilon
(
C
_
1
)
$
. This is ensured by multiplying the coordinates
by
$
\epsilon
$
. More formally, for any vertex
$
v
\in
V
(
G
_
2
)
\setminus
\{
v
_
1
,
\ldots
,v
_
k
\}
$
, we set
$
h
(
v
)[
\sigma
(
i
)]
=
p
(
C
_
1
)[
\sigma
(
i
)]
+
\epsilon
h
_
2
(
v
)[
i
]
$
for
$
i
\in
\{
1
,
\ldots
,d
_
2
\}
$
, and
$
h
(
v
)[
j
]
=
[
p
(
C
_
1
)[
j
]
, p
(
C
_
1
)[
j
]
+
\epsilon
/
2
]
$
, otherwise (for any
$
j
$
not in the
image of
$
\sigma
$
). Note that if we apply the same mapping from
$
h
_
2
$
to
$
h
$
, to the boxes
$
h
_
2
(
v
_
i
)
$
for
$
i
\in
\{
1
,
\ldots
,k
\}
$
, then the
image of
$
h
_
2
(
v
_
i
)
$
fits inside the (previously defined) box
$
h
(
v
_
i
)
$
. Similarly the clique points
$
p
_
2
(
C
)
$
become
$
p
(
C
)
$
by
setting
$
p
(
C
)[
\sigma
(
i
)]
=
p
(
C
_
1
)[
\sigma
(
i
)]
+
\epsilon
p
_
2
(
C
)[
i
]
$
for
$
i
\in
\{
1
,
\ldots
,d
_
2
\}
$
, and
$
p
(
C
)[
j
]
=
p
(
C
_
1
)[
j
]
+
\epsilon
/
4
$
, otherwise.
Note that we have defined (differently) both
$
h
^
\epsilon
(
C
_
1
)
$
(resp.
$
p
(
C
_
1
)
$
) and
$
h
^
\epsilon
(
C
^
*
_
2
)
$
(resp.
$
p
(
C
^
*
_
2
)
$
), despite
the fact that those cliques were merged. In the following we use
$
h
^
\epsilon
(
C
_
1
)
$
and
$
p
(
C
_
1
)
$
only for the purpose of the
proof. The point and the box corresponding to this clique in
$
h
$
is
$
p
(
C
^
*
_
2
)
$
and
$
h
^
\epsilon
(
C
^
*
_
2
)
$
.
Let us now check that
$
h
$
is a
$
C
^
*
_
1
$
-clique sum extendable
vertices of
$
C
^
\star
_
2
$
are also denoted
$
v
_
1
,
\ldots
,v
_
k
$
, and
without loss of generality, the
\textbf
{
(vertices)
}
conditions are
satisfied by setting
$
d
_{
v
_
i
}
=
i
$
for
$
i
\in\{
1
,
\ldots
,k
\}
$
We are now ready to define
$
h
$
. For
$
v
\in
V
(
G
_
1
)
$
, we set
$
h
(
v
)=
h
_
1
(
v
)
$
.
We now scale and translate
$
h
_
2
$
to fit inside
$
h
_
1
^
\varepsilon
(
C
_
1
)
$
.
That is, we fix
$
\varepsilon
>
0
$
small enough so that
\begin{itemize}
\item
the conditions
\textbf
{
(cliques)
}
hold for
$
h
_
1
$
,
\item
$
h
_
1
^
\varepsilon
(
C
_
1
)
\subset
[
0
,
1
)
^
d
$
, and
\item
$
h
_
1
^
\varepsilon
(
C
_
1
)
\sqsubseteq
h
_
1
(
u
)
$
for every
$
u
\in
V
(
G
_
1
)
$
,
\end{itemize}
and for each
$
v
\in
V
(
G
_
2
)
\setminus
V
(
C
^
\star
_
2
)
$
,
we set
$
h
(
v
)[
i
]=
p
_
1
(
C
_
1
)[
i
]
+
\varepsilon
h
_
2
(
v
)[
i
]
$
for
$
i
\in\{
1
,
\ldots
,d
\}
$
.
Note that the condition (v2) for
$
h
_
2
$
implies
$
h
(
v
)
\subset
h
_
1
^
\varepsilon
(
C
_
1
)
$
.
Each clique
$
C
$
of
$
H
$
is a clique of
$
G
_
1
$
or
$
G
_
2
$
.
If
$
C
$
is a clique of
$
G
_
2
$
, we set
$
p
(
C
)=
p
_
1
(
C
_
1
)+
\varepsilon
p
_
2
(
C
)
$
,
otherwise we set
$
p
(
C
)=
p
_
1
(
C
)
$
. In particular, for subcliques of
$
C
_
1
=
C
^
\star
_
2
$
,
we use the former choice.
Let us now check that
$
h
$
is a
$
C
^
\star
_
1
$
-clique sum extendable
representation by comparable boxes. The fact that the boxes are
comparable follows from the fact that those of
$
V
(
G
_
1
)
$
(resp.
$
V
(
G
_
2
)
$
) are comparable in
$
h
_
1
$
(resp.
$
h
_
2
$
) with the
boxes of
$
V
(
C
^
*
_
1
)
$
(resp.
$
V
(
C
^
*
_
2
)
$
) being hypercubes of side one,
and the other boxes being smaller. Clearly, by construction both
$
h
_
1
(
u
)
\sqsubseteq
h
_
1
(
v
)
$
or
$
h
_
2
(
u
)
\sqsubseteq
h
_
2
(
v
)
$
, imply
$
h
(
u
)
\sqsubseteq
h
(
v
)
$
, and for any vertex
$
u
\in
V
(
G
_
1
)
$
and any
vertex
$
v
\in
V
(
G
_
2
)
\setminus
\{
v
_
1
,
\ldots
,v
_
k
\}
$
, we have
$
h
(
v
)
\sqsubseteq
h
^
\epsilon
(
C
_
1
)
\sqsubseteq
h
(
u
)
$
.
comparable follows from the fact that those of
$
h
_
1
$
and
$
h
_
2
$
are comparable and from the scaling of
$
h
_
2
$
: By construction both
$
h
_
1
(
v
)
\sqsubseteq
h
_
1
(
u
)
$
and
$
h
_
2
(
v
)
\sqsubseteq
h
_
2
(
u
)
$
imply
$
h
(
v
)
\sqsubseteq
h
(
u
)
$
, and for any vertex
$
u
\in
V
(
G
_
1
)
$
and any
vertex
$
v
\in
V
(
G
_
2
)
\setminus
V
(
C
^
\star
_
2
)
$
, we have
$
h
(
v
)
\subset
h
_
1
^
\varepsilon
(
C
_
1
)
\sqsubseteq
h
(
u
)
$
.
We now check that
$
h
$
is a contact representation of
$
G
$
. For
$
u,v
\in
V
(
G
_
1
)
$
(resp.
$
u,v
\in
V
(
G
_
2
)
\setminus
\{
v
_
1
,
\ldots
,v
_
k
\}
$
) it
\in
V
(
G
_
1
)
$
(resp.
$
u,v
\in
V
(
G
_
2
)
\setminus
V
(
C
^
\star
_
2
)
$
) it
is clear that
$
h
(
u
)
$
and
$
h
(
v
)
$
have disjoint interiors, and that they
intersect if and only if
$
h
_
1
(
u
)
$
and
$
h
_
1
(
v
)
$
intersect (resp. if
$
h
_
2
(
u
)
$
and
$
h
_
2
(
v
)
$
intersect). Consider now a vertex
$
u
\in
V
(
G
_
1
)
$
and a vertex
$
v
\in
V
(
G
_
2
)
\setminus
\{
v
_
1
,
\ldots
,v
_
k
\}
$
. As
$
h
(
v
)
$
fits inside
$
h
^
\epsilon
(
C
_
1
)
$
, we have that
$
h
(
u
)
$
and
$
h
(
v
)
$
have disjoint interiors. Furthermore, if they intersect then
$
u
\in
V
(
C
_
1
)
$
, say
$
u
=
v
_
1
$
, and
$
h
(
v
)[
1
]
=
[
p
(
C
_
1
)[
1
]
, p
(
C
_
1
)[
1
]+
\alpha
]
$
for some
$
\alpha
>
0
$
. By construction, this implies that
$
h
_
2
(
v
_
1
)
$
and
$
h
_
2
(
v
)
$
intersect.
Finally for the
$
C
^
*
_
1
$
-clique-sum extendability, one can easily
check that the (vertices) conditions hold. For the (cliques)
conditions, as the mapping from
$
p
_
2
(
C
)
$
to
$
p
(
C
)
$
(extended to a
mapping from
$
\mathbb
{
R
}^{
d
_
2
}$
to
$
\mathbb
{
R
}^{
d
}$
) is injective,
we have that (c1) clearly holds. For (c2) one has to notice that if
$
d
=
d
_
2
$
, then the mapping from
$
h
_
2
$
to
$
h
$
extended to the clique boxes
would lead to the same clique boxes
$
h
^{
\epsilon
'
}
(
C
)
$
, with the same point
$
p
(
C
)
$
in their lower corner.
If there are extra dimensions, that is if
$
d> d
_
2
$
, then for any such
dimension
$
j
$
that is not in the image of
$
\sigma
$
, we have that
$
h
^{
\epsilon
'
}
(
C
)[
j
]
=
[
p
(
C
_
1
)[
j
]
+
\epsilon
/
4
, p
(
C
_
1
)[
j
]
+
\epsilon
/
4
+
\epsilon
'
]
\subset
[
p
(
C
_
1
)[
j
]
, p
(
C
_
1
)[
j
]
+
\epsilon
/
2
]
=
h
(
v
)[
j
]
$
.
V
(
G
_
1
)
$
and a vertex
$
v
\in
V
(
G
_
2
)
\setminus
V
(
C
^
\star
_
2
)
$
. As
$
h
(
v
)
\subset
h
^
\varepsilon
(
C
_
1
)
$
, the condition (v2) for
$
h
_
1
$
implies
that
$
h
(
u
)
$
and
$
h
(
v
)
$
have disjoint interiors.
Furthermore, if
$
uv
\in
E
(
G
)
$
, then
$
u
\in
V
(
C
_
1
)=
V
(
C
^
\star
_
2
)
$
, say
$
u
=
v
_
1
$
.
Since
$
uv
\in
E
(
G
_
2
)
$
, the intervals
$
h
_
2
(
u
)[
1
]
$
and
$
h
_
2
(
v
)[
1
]
$
intersect,
and by (v1) and (v2) for
$
h
_
2
$
, we conclude that
$
h
_
2
(
v
)[
1
]=[
0
,
\alpha
]
$
for some positive
$
\alpha
<
1
$
.
Therefore,
$
p
_
1
(
C
_
1
)[
1
]
\in
h
(
v
)[
1
]
$
. Since
$
p
_
1
(
C
_
1
)
\in
\bigcap
_{
x
\in
V
(
C
_
1
)
}
h
_
1
(
x
)
$
,
we have
$
p
_
1
(
C
_
1
)
\in
h
(
u
)
$
, and thus
$
p
_
1
(
C
_
1
)[
1
]
\in
h
(
u
)[
1
]
\cap
h
(
v
)[
1
]
$
.
For
$
i
\in
\{
2
,
\ldots
,d
\}
$
, note that
$
i
\neq
1
=
i
_{
C
_
1
,u
}$
, and thus
by (c2) for
$
h
_
1
$
, we have
$
h
_
1
^
\varepsilon
(
C
_
1
)[
i
]
\subseteq
h
_
1
(
u
)[
i
]=
h
(
u
)[
i
]
$
.
Since
$
h
(
v
)[
i
]
\subseteq
h
_
1
^
\varepsilon
(
C
_
1
)[
i
]
$
, it follows that
$
h
(
u
)
$
intersects
$
h
(
v
)
$
.
Finally, let us consider the
$
C
^
\star
_
1
$
-clique-sum extendability. The
\textbf
{
(vertices)
}
conditions hold, since (v0) and (v1) are inherited from
$
h
_
1
$
, and
(v2) is inherited from
$
h
_
1
$
for
$
v
\in
V
(
G
_
1
)
\setminus
V
(
C
^
\star
_
1
)
$
and follows from the fact that
$
h
(
v
)
\subseteq
h
_
1
^
\varepsilon
(
C
_
1
)
\subset
[
0
,
1
)
^
d
$
for
$
v
\in
V
(
G
_
2
)
\setminus
V
(
C
^
\star
_
2
)
$
. For the
\textbf
{
(cliques)
}
condition (c1),
the mapping
$
p
$
inherits injectivity when restricted to cliques of
$
G
_
2
$
,
or to cliques of
$
G
_
1
$
not contained in
$
C
_
1
$
. For any clique
$
C
$
of
$
G
_
2
$
,
the point
$
p
(
C
)
$
is contained in
$
h
_
1
^
\varepsilon
(
C
_
1
)
$
, since
$
p
_
2
(
C
)
\in
[
0
,
1
)
^
d
$
.
On the other hand, if
$
C'
$
is a clique of
$
G
_
1
$
not contained in
$
C
_
1
$
, then there
exists
$
v
\in
V
(
C'
)
\setminus
V
(
C
_
1
)
$
, we have
$
p
(
C'
)=
p
_
1
(
C'
)
\in
h
_
1
(
v
)
$
, and
$
h
_
1
(
v
)
\cap
h
_
1
^
\varepsilon
(
C
_
1
)=
\emptyset
$
by (c2) for
$
h
_
1
$
.
Therefore, the mapping
$
p
$
is injective, and thus for sufficiently small
$
\varepsilon
'>
0
$
,
we have
$
h
^{
\varepsilon
'
}
(
C
)
\cap
h
^{
\varepsilon
'
}
(
C'
)=
\emptyset
$
for any distinct
cliques
$
C
$
and
$
C'
$
of
$
G
$
.
The condition (c2) of
$
h
$
is (for sufficiently small
$
\varepsilon
'>
0
$
)
inherited from the property (c2) of
$
h
_
1
$
and
$
h
_
2
$
when
$
C
$
is a clique of
$
G
_
2
$
and
$
v
\in
V
(
G
_
2
)
\setminus
V
(
C
^
\star
_
2
)
$
, or
when
$
C
$
is a clique of
$
G
_
1
$
not contained in
$
C
_
1
$
and
$
v
\in
V
(
G
_
1
)
$
.
If
$
C
$
is a clique of
$
G
_
1
$
not contained in
$
C
_
1
$
and
$
v
\in
V
(
G
_
2
)
\setminus
V
(
C
^
\star
_
2
)
$
,
then by (c1) for
$
h
_
1
$
we have
$
h
_
1
^
\varepsilon
(
C
)
\cap
h
_
1
^
\varepsilon
(
C
_
1
)=
\emptyset
$
,
and since
$
h
^{
\varepsilon
'
}
(
C
)
\subseteq
h
_
1
^
\varepsilon
(
C
)
$
and
$
h
(
v
)
\subseteq
h
_
1
^
\varepsilon
(
C
_
1
)
$
,
we conclude that
$
h
(
v
)
\cap
h
^{
\varepsilon
'
}
(
C
)=
\emptyset
$
.
It remains to consider the case that
$
C
$
is a clique of
$
G
_
2
$
and
$
v
\in
V
(
G
_
1
)
$
.
Note that
$
h
^{
\varepsilon
'
}
(
C
)
\subseteq
h
_
1
^
\varepsilon
(
C
_
1
)
$
.
\begin{itemize}
\item
If
$
v
\not\in
V
(
C
_
1
)
$
, then by the property (c2) of
$
h
_
1
$
, the box
$
h
(
v
)=
h
_
1
(
v
)
$
is disjoint from
$
h
_
1
^
\varepsilon
(
C
_
1
)
$
,
and thus
$
h
(
v
)
\cap
h
^{
\varepsilon
'
}
(
C
)=
\emptyset
$
.
\item
Otherwise
$
v
\in
V
(
C
_
1
)=
V
(
C
^
\star
_
2
)
$
, say
$
v
=
v
_
1
$
.
Note that by (v1), we have
$
h
_
2
(
v
)=[-
1
,
0
]
\times
[
0
,
1
]
^{
d
-
1
}$
.
\begin{itemize}
\item
If
$
v
\not\in
V
(
C
)
$
, then by the property (c2) of
$
h
_
2
$
, the box
$
h
_
2
(
v
)
$
is disjoint from
$
h
_
2
^
\varepsilon
(
C
)
$
.
Since
$
h
_
2
^
\varepsilon
(
C
)[
i
]
\subseteq
[
0
,
1
]=
h
_
2
(
v
)[
i
]
$
for
$
i
\in\{
2
,
\ldots
,d
\}
$
,
it follows that
$
h
_
2
^
\varepsilon
(
C
)[
1
]
\subseteq
(
0
,
1
)
$
, and thus
$
h
^{
\varepsilon
'
}
(
C
)[
1
]
\subseteq
h
_
1
^
\varepsilon
(
C
_
1
)[
1
]
\setminus\{
p
(
C
_
1
)[
1
]
\}
$
.
By (c2) for
$
h
_
1
$
, we have
$
h
(
v
)[
1
]
\cap
h
_
1
^
\varepsilon
(
C
_
1
)[
1
]=
h
_
1
(
v
)[
1
]
\cap
h
_
1
^
\varepsilon
(
C
_
1
)[
1
]=
p
(
C
_
1
)[
1
]
$
,
and thus
$
h
(
v
)
\cap
h
^{
\varepsilon
'
}
(
C
)=
\emptyset
$
.
\item
If
$
v
\in
V
(
C
)
$
, then by the property (c2) of
$
h
_
2
$
, the intersection of
$
h
_
2
(
v
)[
1
]=[-
1
,
0
]
$
and
$
h
_
2
^
\varepsilon
(
C
)[
1
]
\subseteq
[
0
,
1
)
$
is the single point
$
p
_
2
(
C
)[
1
]=
0
$
,
and thus
$
p
(
C
)[
1
]=
p
_
1
(
C
_
1
)[
1
]
$
and
$
h
^{
\varepsilon
'
}
(
C
)[
1
]=[
p
_
1
(
C
_
1
)[
1
]
,p
_
1
(
C
_
1
)[
1
]+
\varepsilon
'
]
$
.
Recall that the property (c2) of
$
h
_
1
$
implies
$
h
(
v
)[
1
]
\cap
h
_
1
^
\varepsilon
(
C
_
1
)[
1
]=
\{
p
(
C
_
1
)[
1
]
\}
$
,
and thus
$
h
(
v
)[
1
]
\cap
h
^{
\varepsilon
'
}
(
C
)[
1
]=
\{
p
(
C
)[
1
]
\}
$
. For
$
i
\in\{
2
,
\ldots
, d
\}
$
,
the property (c2) of
$
h
_
1
$
implies
$
h
_
1
^
\varepsilon
(
C
_
1
)[
i
]
\subseteq
h
_
1
(
v
)[
i
]=
h
(
v
)[
i
]
$
, and
since
$
h
^{
\varepsilon
'
}
(
C
)[
i
]
\subseteq
h
_
1
^
\varepsilon
(
C
_
1
)[
i
]
$
, it follows that
$
h
^{
\varepsilon
'
}
(
C
)[
i
]
\subseteq
h
(
v
)[
i
]
$
.
\end{itemize}
\end{itemize}
\end{proof}
The following lemma shows that any graphs has a
$
C
^
*
$
-clique-sum
The following lemma shows that any graphs has a
$
C
^
\star
$
-clique-sum
extendable representation in
$
\mathbb
{
R
}^
d
$
, for
$
d
=
\omega
(
G
)
+
\ecbdim
(
G
)
$
and for any clique
$
C
^
*
$
.
\ecbdim
(
G
)
$
and for any clique
$
C
^
\star
$
.
\begin{lemma}
\label
{
lem-apex-cs
}
For any graph
$
G
$
and any clique
$
C
^
*
$
, we have that
$
G
$
admits a
$
C
^
*
$
-clique-sum extendable touching representation by comparabe
boxes in
$
\mathbb
{
R
}^
d
$
, for
$
d
=
|V
(
C
^
*
)
|
+
\ecbdim
(
G
\setminus
V
(
C
^
*
))
$
.
For any graph
$
G
$
and any clique
$
C
^
\star
$
, we have that
$
G
$
admits a
$
C
^
\star
$
-clique-sum extendable touching representation by comparabe
boxes in
$
\mathbb
{
R
}^
d
$
, for
$
d
=
|V
(
C
^
\star
)
|
+
\ecbdim
(
G
\setminus
V
(
C
^
\star
))
$
.
\end{lemma}
\begin{proof}
The proof is essentially the same as the one of
Lemma~
\ref
{
lemma-apex
}
. Consider a
$
\emptyset
$
-clique-sum
extendable touching representation
$
h'
$
of
$
G
\setminus
V
(
C
^
*
)
$
by
extendable touching representation
$
h'
$
of
$
G
\setminus
V
(
C
^
\star
)
$
by
comparable boxes in
$
\mathbb
{
R
}^{
d'
}$
, with
$
d'
=
\cbdim
(
G
\setminus
V
(
C
^
*
))
$
, and let
$
V
(
C
^
*
)
=
\{
v
_
1
,
\ldots
,v
_
k
\}
$
. We now construct
V
(
C
^
\star
))
$
, and let
$
V
(
C
^
\star
)
=
\{
v
_
1
,
\ldots
,v
_
k
\}
$
. We now construct
the desired representation
$
h
$
of
$
G
$
as follows. For each vertex
$
v
_
i
\in
V
(
C
^
*
)
$
let
$
h
(
v
_
i
)
$
be the box fulfilling (v1) with
$
d
_{
v
_
i
}
=
i
$
. For each vertex
$
u
\in
V
(
G
)
\setminus
V
(
C
^
*
)
$
, if
$
i
\le
$
v
_
i
\in
V
(
C
^
\star
)
$
let
$
h
(
v
_
i
)
$
be the box fulfilling (v1) with
$
d
_{
v
_
i
}
=
i
$
. For each vertex
$
u
\in
V
(
G
)
\setminus
V
(
C
^
\star
)
$
, if
$
i
\le
k
$
then let
$
h
(
u
)[
i
]
=
[
0
,
1
/
2
]
$
if
$
uv
_
i
\in
E
(
G
)
$
, and
$
h
(
u
)[
i
]
=
[
1
/
4
,
3
/
4
]
$
if
$
uv
_
i
\notin
E
(
G
)
$
. For
$
i>k
$
we have
$
h
(
u
)[
i
]
=
\alpha
_
i h'
(
u
)[
i
-
k
]
$
, for some
$
\alpha
_
i>
0
$
. The values
$
\alpha
_
i>
0
$
are chosen suffciently small so that
$
h
(
u
)[
i
]
\subset
[
0
,
1
)
$
, whenever
$
u
\notin
V
(
C
^
*
)
$
.
are chosen suffciently small so that
$
h
(
u
)[
i
]
\subset
[
0
,
1
)
$
, whenever
$
u
\notin
V
(
C
^
\star
)
$
.
We proceed similarly for the clique points. For any
clique
$
C
$
of
$
G
$
, if
$
i
\le
k
$
then let
$
p
(
C
)[
i
]
=
0
$
if
$
v
_
i
\in
V
(
C
)
$
, and
$
p
(
C
)[
i
]
=
1
/
4
$
if
$
v
_
i
\notin
V
(
C
)
$
. For
$
i>k
$
we have
...
...
@@ -487,30 +545,30 @@ extendable representation in $\mathbb{R}^d$, for $d= \omega(G) +
\{
v
_
1
,
\ldots
,v
_
k
\}
$
, as we set
$
p
(
C
)[
i
]
=
\alpha
_
i p'
(
C'
)[
i
-
k
]
$
.
As
$
h'
(
u
)
\sqsubset
h'
(
v
)
$
implies that
$
h
(
u
)
\sqsubset
h
(
v
)
$
, and as
$
h
(
u
)
\sqsubset
h
(
v
_
i
)
$
, for every
$
u
\in
V
(
G
)
\setminus
V
(
C
^
*
)
$
and every
$
v
_
i
\in
V
(
C
^
*
)
$
, we have that
$
h
$
is a touching representation by comparable boxes.
$
h
(
u
)
\sqsubset
h
(
v
_
i
)
$
, for every
$
u
\in
V
(
G
)
\setminus
V
(
C
^
\star
)
$
and every
$
v
_
i
\in
V
(
C
^
\star
)
$
, we have that
$
h
$
is a touching representation by comparable boxes.
By the construction, it is clear that
$
h
$
is a representation of
$
G
$
.
For the
$
C
^
*
$
-clique-sum extendability, it is clear that the (vertices) conditions hold.
For the
$
C
^
\star
$
-clique-sum extendability, it is clear that the (vertices) conditions hold.
For the (cliques) condition (c1), let us first consider two distinct cliques
$
C
_
1
$
and
$
C
_
2
$
of
$
G
$
such that
$
|V
(
C
_
1
)
|
\ge
|V
(
C
_
2
)
|
$
, and let
$
C'
_
i
=
C
_
i
\setminus
V
(
C
^
*
)
$
. If
$
C'
_
1
=
C'
_
2
$
,
of
$
G
$
such that
$
|V
(
C
_
1
)
|
\ge
|V
(
C
_
2
)
|
$
, and let
$
C'
_
i
=
C
_
i
\setminus
V
(
C
^
\star
)
$
. If
$
C'
_
1
=
C'
_
2
$
,
there is a vertex
$
v
_
i
\in
V
(
C
_
1
)
\setminus
V
(
C
_
2
)
$
, and
$
p
(
C
_
1
)[
i
]
=
0
\neq
1
/
4
=
p
(
C
_
2
)[
i
]
$
.
Otherwise, if
$
C'
_
1
\neq
C'
_
2
$
, we have that
$
p'
(
C'
_
1
)
\neq
p'
(
C'
_
2
)
$
, which leads to
$
p
(
C
_
1
)
\neq
p
(
C
_
2
)
$
by construction.
For the (cliques) condition (c2), let us first consider a vertex
$
v
\in
V
(
G
)
\setminus
V
(
C
^
*
)
$
and a clique
$
C
$
of
$
G
$
containing
$
v
$
.
In the first dimensions
$
i
\le
k
$
, we always have
$
h
^
\epsilon
(
C
)[
i
]
\subseteq
h
(
v
)[
i
]
$
. Indeed, if
$
v
_
i
\in
V
(
C
)
$
we have
$
h
^
\epsilon
(
C
)[
i
]
\subseteq
[
0
,
1
/
2
]
=
h
(
v
)[
i
]
$
(as in that case
$
v
$
and
$
v
_
i
$
are adjacent), and if
$
v
_
i
\notin
V
(
C
)
$
we have
$
h
^
\epsilon
(
C
)[
i
]
\subseteq
[
1
/
4
,
1
/
2
]
\subseteq
h
(
v
)[
i
]
$
. Then for the last
$
d'
$
dimensions, by definition of
$
h'
$
,
we have that
$
h
^
\epsilon
(
C
)[
i
]
\subseteq
h
(
v
)[
i
]
$
for every
$
i>k
$
, except one,
for which
$
h
^
\epsilon
(
C
)[
i
]
\cap
h
(
v
)[
i
]
=
\{
p
(
C
)[
i
]
\}
$
. This completes the first case
and we now consider a vertex
$
v
\in
V
(
G
)
\setminus
V
(
C
^
*
)
$
and a clique
$
C
$
of
$
G
$
not containing
$
v
$
.
For the (cliques) condition (c2), let us first consider a vertex
$
v
\in
V
(
G
)
\setminus
V
(
C
^
\star
)
$
and a clique
$
C
$
of
$
G
$
containing
$
v
$
.
In the first dimensions
$
i
\le
k
$
, we always have
$
h
^
\
var
epsilon
(
C
)[
i
]
\subseteq
h
(
v
)[
i
]
$
. Indeed, if
$
v
_
i
\in
V
(
C
)
$
we have
$
h
^
\
var
epsilon
(
C
)[
i
]
\subseteq
[
0
,
1
/
2
]
=
h
(
v
)[
i
]
$
(as in that case
$
v
$
and
$
v
_
i
$
are adjacent), and if
$
v
_
i
\notin
V
(
C
)
$
we have
$
h
^
\
var
epsilon
(
C
)[
i
]
\subseteq
[
1
/
4
,
1
/
2
]
\subseteq
h
(
v
)[
i
]
$
. Then for the last
$
d'
$
dimensions, by definition of
$
h'
$
,
we have that
$
h
^
\
var
epsilon
(
C
)[
i
]
\subseteq
h
(
v
)[
i
]
$
for every
$
i>k
$
, except one,
for which
$
h
^
\
var
epsilon
(
C
)[
i
]
\cap
h
(
v
)[
i
]
=
\{
p
(
C
)[
i
]
\}
$
. This completes the first case
and we now consider a vertex
$
v
\in
V
(
G
)
\setminus
V
(
C
^
\star
)
$
and a clique
$
C
$
of
$
G
$
not containing
$
v
$
.
As
$
v
\notin
V
(
C'
)
$
, there is an hyperplane
${
\mathcal
H
}
'
=
\{
p
\in
\mathbb
{
R
}^{
d'
}
\
|
\
p
[
i
]
=
c
\}
$
that separates
$
p'
(
C'
)
$
and
$
h'
(
v
)
$
.
This implies that the following hyperplane
${
\mathcal
H
}
=
\{
p
\in
\mathbb
{
R
}^{
d
}
\
|
\
p
[
k
+
i
]
=
\alpha
_{
k
+
i
}
c
\}
$
separates
$
p
(
C
)
$
and
$
h
(
v
)
$
.
Now we consider a vertex
$
v
_
i
\in
V
(
C
^
*
)
$
, and we note that for any clique
$
C
$
containing
$
v
_
i
$
we have that
$
h
^
\epsilon
(
C
)[
i
]
\cap
h
(
v
_
i
)[
i
]
=
[
0
,
\epsilon
]
\cap
[-
1
,
0
]
=
\{
0
\}
$
, and
$
h
^
\epsilon
(
C
)[
j
]
\subseteq
[
0
,
1
]
=
h
(
v
_
i
)[
j
]
$
Now we consider a vertex
$
v
_
i
\in
V
(
C
^
\star
)
$
, and we note that for any clique
$
C
$
containing
$
v
_
i
$
we have that
$
h
^
\
var
epsilon
(
C
)[
i
]
\cap
h
(
v
_
i
)[
i
]
=
[
0
,
\
var
epsilon
]
\cap
[-
1
,
0
]
=
\{
0
\}
$
, and
$
h
^
\
var
epsilon
(
C
)[
j
]
\subseteq
[
0
,
1
]
=
h
(
v
_
i
)[
j
]
$
for any
$
j
\neq
i
$
. For a clique
$
C
$
that does not contain
$
v
_
i
$
we have that
$
h
^
\epsilon
(
C
)[
i
]
\cap
h
(
v
_
i
)[
i
]
\subset
(
0
,
1
)
\cap
[-
1
,
0
]
=
\emptyset
$
.
$
h
^
\
var
epsilon
(
C
)[
i
]
\cap
h
(
v
_
i
)[
i
]
\subset
(
0
,
1
)
\cap
[-
1
,
0
]
=
\emptyset
$
.
Condition (c2) is thus fulfilled and this completes the proof of the lemma.
\end{proof}
...
...
@@ -526,7 +584,7 @@ of $\cbdim(G)$ and $\chi(G)$.
of
$
h
$
. We first scale
$
h
$
to fit in
$
(
0
,
1
)
^
d
$
, and for a
sufficiently small real
$
\alpha
>
0
$
we increase each box in
$
h
$
, by
$
2
\alpha
$
in every dimension, that is we replace
$
h
(
v
)[
i
]
=
[
a,b
]
$
by
$
[
a
-
\epsilon
,b
+
\epsilon
]
$
for each vertex
$
v
$
and dimension
by
$
[
a
-
\
var
epsilon
,b
+
\
var
epsilon
]
$
for each vertex
$
v
$
and dimension
$
i
$
. Furthermore
$
\alpha
$
is chosen sufficiently small, so that no
new intersection was created. The obtained representation
$
h
_
1
$
is
thus an intersection representation of the same graph
$
G
$
such that,
...
...
@@ -536,7 +594,7 @@ of $\cbdim(G)$ and $\chi(G)$.
Now we add
$
\chi
(
G
)
$
dimensions to make the representation touching
again, and to ensure some space for the clique boxes
$
h
^
\epsilon
(
C
)
$
. Formally we define
$
h
_
2
$
as follows.
$
h
^
\
var
epsilon
(
C
)
$
. Formally we define
$
h
_
2
$
as follows.
$$
h
_
2
(
u
)[
i
]=
\begin
{
cases
}
h
_
1
(
u
)[
i
]
&
\text
{
if $i
\le
d$
}
\\
[
1
/
5
,
3
/
5
]
&
\text
{
if $c
(
u
)
< i
-
d$
}
\\
...
...
@@ -566,13 +624,13 @@ of $\cbdim(G)$ and $\chi(G)$.
which clique points
$
p
_
2
(
C
_
1
)
$
and
$
p
_
2
(
C
_
2
)
$
are based on distinct maximum cliques, necessarily lead to distinct points.
In the case that
$
C
_
1
$
and
$
C
_
2
$
belong to some maximal clique
$
C
$
, we have that
$
c
(
C
_
1
)
\neq
c
(
C
_
2
)
$
and this implies by construction that
$
p
_
2
(
C
_
1
)
$
and
$
p
_
2
(
C
_
2
)
$
are distinct. Thus (c1) holds.
By construction of
$
h
_
1
$
, we have that if
$
h
_
2
^{
\epsilon
}
(
C'
)[
i
]
\cap
h
_
2
(
v
)[
i
]
$
is non-empty for every
$
i
\le
d
$
,
then we have that
$
h
_
2
^{
\epsilon
}
(
C'
)[
i
]
\subset
h
_
2
(
v
)[
i
]
$
for every
$
i
\le
d
$
,
By construction of
$
h
_
1
$
, we have that if
$
h
_
2
^{
\
var
epsilon
}
(
C'
)[
i
]
\cap
h
_
2
(
v
)[
i
]
$
is non-empty for every
$
i
\le
d
$
,
then we have that
$
h
_
2
^{
\
var
epsilon
}
(
C'
)[
i
]
\subset
h
_
2
(
v
)[
i
]
$
for every
$
i
\le
d
$
,
and we have that
$
v
$
belongs to some maximal clique
$
C
$
containing
$
C'
$
. If
$
v
\notin
V
(
C'
)
$
note that
$
p
_
2
(
C'
)[
d
+
c
(
v
)]
=
1
/
2
\notin
[
0
,
2
/
5
]=
h
_
2
(
v
)[
d
+
c
(
v
)]
$
, while if
$
v
\in
V
(
C'
)
$
we have that
$
h
_
2
^{
\epsilon
}
(
C'
)[
i
]
\subset
[
2
/
5
,
1
/
2
+
\epsilon
]
\subset
h
_
2
(
v
)[
i
]
$
for every dimension
$
i>d
$
,
except if
$
c
(
v
)=
i
-
d
$
, and in that case
$
h
_
2
(
v
)[
i
]
\cap
h
_
2
^{
\epsilon
}
(
C'
)[
i
]
=
[
0
,
2
/
5
]
\cap
[
2
/
5
,
2
/
5
+
\epsilon
]
=
\{
2
/
5
\}
$
. We thus have that (c2) holds, and this concludes the proof of the lemma.
$
h
_
2
^{
\
var
epsilon
}
(
C'
)[
i
]
\subset
[
2
/
5
,
1
/
2
+
\
var
epsilon
]
\subset
h
_
2
(
v
)[
i
]
$
for every dimension
$
i>d
$
,
except if
$
c
(
v
)=
i
-
d
$
, and in that case
$
h
_
2
(
v
)[
i
]
\cap
h
_
2
^{
\
var
epsilon
}
(
C'
)[
i
]
=
[
0
,
2
/
5
]
\cap
[
2
/
5
,
2
/
5
+
\
var
epsilon
]
=
\{
2
/
5
\}
$
. We thus have that (c2) holds, and this concludes the proof of the lemma.
\end{proof}
...
...
@@ -605,14 +663,14 @@ in $\mathbb{R}^{\lceil \log_2 m \rceil}$ respectively, by
For any
$
k
$
-tree
$
G
$
,
$
\cbdim
(
G
)
\le
\ecbdim
(
G
)
\le
k
+
1
$
.
\end{theorem}
\begin{proof}
Note that there exists a
$
k
$
-tree
$
G'
$
having a
$
k
$
-clique
$
C
^
*
$
such that
$
G'
\setminus
V
(
C
^
*
)
$
corresponds to
$
G
$
. Let us construct
a
$
C
^
*
$
-clique-sum extendable representation of
$
G'
$
and note that
Note that there exists a
$
k
$
-tree
$
G'
$
having a
$
k
$
-clique
$
C
^
\star
$
such that
$
G'
\setminus
V
(
C
^
\star
)
$
corresponds to
$
G
$
. Let us construct
a
$
C
^
\star
$
-clique-sum extendable representation of
$
G'
$
and note that
it induces a
$
\emptyset
$
-clique-sum extendable representation of
$
G
$
.
Note that
$
G'
$
can be obtained by starting with a
$
(
k
+
1
)
$
-clique
containing
$
C
^
*
$
, and by performing successive full clique-sums of
containing
$
C
^
\star
$
, and by performing successive full clique-sums of
$
K
_{
k
+
1
}$
on a
$
K
_
k
$
subclique. By Lemma~
\ref
{
lem-cs
}
, it suffice to
show that
$
K
_{
k
+
1
}$
, the
$
(
k
+
1
)
$
-clique with vertex set
$
\{
v
_
1
,
\ldots
, v
_{
k
+
1
}
\}
$
, has a
$
(
K
_{
k
+
1
}
-
\{
v
_{
k
+
1
}
\}
)
$
-clique-sum
...
...
@@ -641,15 +699,15 @@ vertex $v_i\in V(C_1)\setminus V(C_2)$, and this implies that
$
p
(
C
_
1
)[
i
]
< p
(
C
_
2
)[
i
]
$
.
For a vertex
$
v
_
i
$
and a clique
$
C
$
, the boxes
$
h
(
v
_
i
)
$
and
$
h
^
\epsilon
(
C
)
$
intersect if and only if
$
v
_
i
\in
V
(
C
)
$
. Indeed, if
$
v
_
i
\in
V
(
C
)
$
then
$
p
(
C
)
\in
h
(
v
_
i
)
$
and
$
p
(
C
)
\in
h
^
\epsilon
(
C
)
$
, and
$
h
^
\
var
epsilon
(
C
)
$
intersect if and only if
$
v
_
i
\in
V
(
C
)
$
. Indeed, if
$
v
_
i
\in
V
(
C
)
$
then
$
p
(
C
)
\in
h
(
v
_
i
)
$
and
$
p
(
C
)
\in
h
^
\
var
epsilon
(
C
)
$
, and
if
$
v
_
i
\notin
V
(
C
)
$
then
$
h
(
v
_
i
)[
i
]
=
[-
1
,
0
]
$
if
$
i
\le
k
$
(resp.
$
h
(
v
_
i
)[
i
]
=
[
0
,
\frac
12
]
$
if
$
i
=
k
+
1
$
) and
$
h
^
\epsilon
(
C
)[
i
]
=
[
\frac
14
,
\frac
14
+
\epsilon
]
$
(resp.
$
h
^
\epsilon
(
C
)[
i
]
=
[
\frac
34
,
\frac
34
+
\epsilon
]
$
). Finally, if
$
v
_
i
\in
V
(
C
)
$
we have that
$
h
(
v
_
i
)[
i
]
\cap
h
^
\epsilon
(
C
)[
i
]
=
\{
p
(
C
)[
i
]
\}
$
and that
$
h
(
v
_
i
)[
j
]
\cap
h
^
\epsilon
(
C
)[
j
]
=
[
p
(
C
)[
j
]
,p
(
C
)[
j
]+
\epsilon
]
$
for any
$
j
\neq
i
$
and any
$
\epsilon
<
\frac
14
$
. This concludes the proof of the theorem.
(resp.
$
h
(
v
_
i
)[
i
]
=
[
0
,
\frac
12
]
$
if
$
i
=
k
+
1
$
) and
$
h
^
\
var
epsilon
(
C
)[
i
]
=
[
\frac
14
,
\frac
14
+
\
var
epsilon
]
$
(resp.
$
h
^
\
var
epsilon
(
C
)[
i
]
=
[
\frac
34
,
\frac
34
+
\
var
epsilon
]
$
). Finally, if
$
v
_
i
\in
V
(
C
)
$
we have that
$
h
(
v
_
i
)[
i
]
\cap
h
^
\
var
epsilon
(
C
)[
i
]
=
\{
p
(
C
)[
i
]
\}
$
and that
$
h
(
v
_
i
)[
j
]
\cap
h
^
\
var
epsilon
(
C
)[
j
]
=
[
p
(
C
)[
j
]
,p
(
C
)[
j
]+
\
var
epsilon
]
$
for any
$
j
\neq
i
$
and any
$
\
var
epsilon
<
\frac
14
$
. This concludes the proof of the theorem.
\end{proof}
Note that actually the bound on the comparable boxes dimension of Theorem~
\ref
{
thm-ktree
}
extends to graphs of treewidth
$
k
$
. For this, note that the construction in this proof can
...
...
@@ -659,7 +717,7 @@ with $h(u) \sqsubset h(v)$ say, the intersection $I = h(u) \cap h(v)$ is a facet
Actually
$
I
[
i
]
=
h
(
u
)[
i
]
$
for every dimension, except one that we denote
$
j
$
. For this
dimension we have that
$
I
[
j
]=
\{
c
\}
$
for some
$
c
$
, and that
$
h
(
u
)[
j
]=[
c,c
+
s
]
$
,
where
$
s
$
is the length of the sides of
$
h
(
u
)
$
. In that context to delete an edge
$
uv
$
one can simply replace
$
h
(
u
)[
j
]=[
c,c
+
s
]
$
with
$
[
c
+
\epsilon
,c
+
s
]
$
, for a sufficiently small
$
\epsilon
$
.
one can simply replace
$
h
(
u
)[
j
]=[
c,c
+
s
]
$
with
$
[
c
+
\
var
epsilon
,c
+
s
]
$
, for a sufficiently small
$
\
var
epsilon
$
.
One can proceed similarly for any subset of edges, and note that as the hypercubes in
$
h
$
have
distinct sizes these small perturbations give rise to boxes that are still comparable.
Thus for any treewidth
$
k
$
graph
$
H
$
(that is a subgraph of a
$
k
$
-tree
$
G
$
) we have
$
\cbdim
(
H
)
\le
k
+
1
$
.
...
...
@@ -689,7 +747,7 @@ and Lemma~\ref{lem-ecbdim-cbdim}, these graphs admit a
$
\emptyset
$
-clique-sum extendable representations in bounded
dimensions. As the obtained graphs have bounded dimension, by
Lemma~
\ref
{
lemma-cliq
}
and Lemma~
\ref
{
lem-apex-cs
}
, for any choice of
a root clique
$
C
^
*
$
, they have a
$
C
^
*
$
-clique-sum extendable
a root clique
$
C
^
\star
$
, they have a
$
C
^
\star
$
-clique-sum extendable
representation in bounded dimension. Thus by Lemma~
\ref
{
lem-cs
}
any
sequence of clique sum from these graphs leads to a graph with bounded
dimension. Finally, we have seen that taking a subgraph does not lead
...
...
This diff is collapsed.
Click to expand it.
Preview
0%
Loading
Try again
or
attach a new file
.
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Save comment
Cancel
Please
register
or
sign in
to comment