2000 Mathematics Subject Classification. Primary 46L87; Secondary 46L07, 53C23, 58B34, 60B10.
Quantized Gromov-Hausdorff distance
wei wu
Department of Mathematics, East China Normal University, Shanghai 200062, P.R. China E-mail address : wwu@math.ecnu.edu.cn Current address : Department of Mathematics, University of California, Berkeley, CA 94720-3840 E-mail address : wwu@math.berkeley.edu
-
Abstract.
A quantized metric space is a matrix order unit space equipped with an operator space version of Rieffel's Lip-norm. We develop for quantized metric spaces an operator space version of quantum Gromov-Hausdorff distance.
We show that two quantized metric spaces are completely isometric if and only if their quantized Gromov-Hausdorff distance is zero. We establish a completeness theorem. As applications, we show that a quantized metric space with 1-exact underlying matrix order unit space is a limit of matrix algebras with respect to quantized Gromov-Hausdorff distance, and that matrix algebras converge naturally to the sphere for quantized Gromov-Hausdorff distance.
1 Introduction
Following up the compact metric spaces given by Connes in connection with his theory of quantum Riemannian geometry defined by Dirac operators [3] , Rieffel defined the notion of a compact quantum metric space
in [17] as an order unit space
equipped with a Lip-norm
, which is a generalization of the usual Lipschitz seminorm on functions which one associates to an ordinary metric. Many interesting examples of compact quantum metric space have been constructed [14, 16, 11, 10] . Motivated by the type of convergence of spaces that has recently begun to play a central role in string theory, Rieffel introduces the quantum Gromov-Hausdorff distance for the compact quantum metric spaces as a quantum analogue of Gromov-Hausdorff distance, and shows that the basic theorems of the classical theory have natural quantum analogues.
In [20] and [21] , we formulated matrix Lipschitz seminorms on matrix order unit spaces. This operator space version of Lipschitz seminorm has many nice properties which are similar to those for ordinary metric spaces. These data may then be thought of as some `noncommutative metric spaces'. So it is natural to ask, as does Rieffel in [17] , if it is possible to develop a corresponding operator space version of quantum Gromov-Hausdorff distance. This is the aim of the present article.
In contrast to the matricial quantum Gromov-Hausdorff distance in [8] and operator Gromov-Hausdorff distance in [9] , our quantized Gromov-Hausdorff distance operates entirely at the “matrix” level. Not only the matrix state spaces but also the matrix Lipschitz seminorms and the complete isometries are brought into our picture. This should be important in the background of operator systems.
The paper has eight sections. Section 2 contains preliminaries, mainly to fix some terminology and notation. In Section 3 we define quantized metric space and develop an operator “quotient”. Section 4 defines our quantized Gromov-Hausdorff distance, and we prove that it satisfies the triangle inequality. Section 5 deals with the operator Gromov-Hausdorff distance zero. We establish that it implies a complete isometry. Section 6 treats the completeness theorem of the complete isometry classes of quantized metric spaces. In Section 7 we show that a quantized metric space with 1-exact underlying matrix order unit space is a limit of matrix algebras with respect to quantized Gromov-Hausdorff distance. It is established in Section 8 that matrix algebras converge naturally to the sphere for quantized Gromov-Hausdorff distance.
2 Preliminaries
All vector spaces are assumed to be complex throughout this paper. Given a vector space
, we let
denote the matrix space of all
by
matrices
with
, and we set
. If
, we write
and
, which means that we may identify
with the tensor product
. We identify
with the normed space
.
We use the standard matrix multiplication and *-operation for compatible scalar matrices, and
for the identity matrix in
.
There are two natural operations on the matrix spaces. For
and
, the direct sum
is defined by letting
and if we are given
,
and
, the matrix product
is defined by
A *-vector space
is a complex vector space together with a conjugate linear mapping
such that
. A complex vector space
is said to be matrix ordered if:
-
(1)
is a *-vector space;
-
(2)
each
,
, is partially ordered;
-
(3)
if
is any
matrix of complex numbers.
A matrix order unit space
is a matrix ordered space
together with a distinguished order unit
satisfying the following conditions:
-
(1)
is a proper cone with the order unit
;
-
(2)
each of the cones
is Archimedean.
Each matrix order unit space
may be provided with the norm
As in [17] , we will not assume that
is complete for the norm.
If
and
are *-vector spaces and
is a linear mapping, we have a linear mapping
defined by
.
Given vector spaces
and
and a linear mapping
and
, we have a corresponding
defined by
If
and
are vector spaces in duality, then they determine the matrix pairing
where
for
and
.
A graded set
is a sequence of sets
. If
is a locally convex topological vector space, then the canonical topology on
is that determined by the natural linear isomorphism
, that is, the product topology. A graded set
with
is closed or compact if that is the case for each set
in the product topology in
. Given a vector space
, we say that a graded set
with
is absolutely matrix convex if for all
-
(1)
;
-
(2)
for any contractions
and
.
A matrix convex set in
is a graded set
of subsets
such that
for all
and
for
satisfying
.
Let
and
be vector spaces in duality, and let
be a graded set with
. The absolute operator polar
with
, is defined by
. The matrix polar
with
, is defined by
. Given a subset
, the absolute polar of
is defined by
.
A gauge on a vector space
is a function
such that
-
(1)
;
-
(2)
,
for all
and
. We say that a gauge
is a seminorm on
if
for all
. Given an arbitrary vector space
, a matrix gauge
on
is a sequence of gauges
such that
-
(1)
;
-
(2)
,
for any
,
,
and
. A matrix gauge
is a matrix seminorm on
if for any
for all
.
If each
is a norm on
, we say that
is a matrix norm. An operator space is a vector space together with a matrix norm on it. For a matrix order unit space
, it is an operator space with the matrix norm determined by the matrix order on it.
3 Quantized metric space
First we recall the following definitions given in [20, 21] :
Definition 3.1.
Given a matrix order unit space
, a matrix Lipschitz seminorm
on
is a sequence of seminorms
such that
-
(1)
the null space of each
is
;
-
(2)
;
-
(3)
;
-
(4)
,
for any
,
,
and
.
Let
be a matrix order unit space. The matrix state space of
is the collection
of matrix states
If
is a matrix Lipschitz seminorm on
, we have a collection
of metrics on
given by
for
(notice that it may take value
). And in turn we obtain a sequence
of gauges on
by
for all
.
Definition 3.2.
Let
be a matrix order unit space. By a matrix Lip-norm on
we mean a matrix Lipschitz seminorm
on
such that the
-topology on
agrees with the BW-topology.
We are now prepared to make:
Definition 3.3.
By a quantized metric space we mean a pair
consisting of a matrix order unit space
with a matrix Lip-norm
defined on it.
Example 3.4.
Let
be an ordinary compact metric space, let
denote the set of Lipschitz functions on
, and let
denote the Lipschitz seminorm on
. Then
, and for
and
, we have
Thus
is a self-adjoint linear subspace of
which contains constant functions, and so
is a matrix order unit space by Theorem 4.4 in [
2]
.
Since
is lower semicontinuous,
is an absolutely convex normed-closed (and hence is weakly closed) set in
.
determines a graded set
The minimal envelope
of
is the matrix bipolar
of
.
is an absolutely matrix convex weakly closed graded set. We let
be the corresponding matrix gauge of
. Since
is a Lipschitz seminorm,
is a matrix Lipschitz seminorm.
implies that
is also a matrix Lip-norm(see Theorem 1.9 in [
14]
and Proposition 7.5 in [
21]
). Therefore,
is a quantized metric space. It is called the minimal quantized metric space of
. The maximal envelope
of
is the matrix polar
of
, where
Similarly,
is an absolutely matrix convex weakly closed graded set, and the corresponding matrix gauge
of
makes
into a quantized metric space.
is called the maximal quantized metric space of
.
Moreover, if
is an absolutely matrix convex weakly closed graded set with
, then
and the corresponding matrix gauge
satisfies
(see page 181 in [
4]
). So
is a quantized metric space. It is called a quantized metric space of
.
Example 3.5.
Let
be a compact quantum metric space, that is, an order unit space
equipped with a seminorm
, called Lip-norm, on
such that
if and only if
, and the topology on the state space
of
from the metric
is the
-topology (see Definition 2.2 in [
17]
). So
is an ordinary compact metric space. Let
denote the set of Lipschitz functions on
. By Example 3.4 , there exists a quantized metric space structure
of
, where
. From Lemma 3.2 in [
15]
,
and
for
. Let
be the matrix norm determined by the matrix order on
. By the basic representation theorem of Kadison[
7]
, we also have that
for
. If
is lower semicontinuous, the embedding of
into
is isometric, that is,
and
for all
, according to Theorem 4.1 in [
15]
.
Set
We denote the restriction of
on
by
. Then
is a self-adjoint linear subspace of
and contains the order unit of
. So
is a matrix order unit space. Because the
-topology on
agrees with the BW-topology, the image of
in
is totally bounded for
by Theorem 5.3 in [
20]
. Since
, the image of
in
is totally bounded for
, and so, by Theorem 5.3 in [
20]
, the
-topology on
is the BW-topology. Therefore,
is a quantized metric space, and the embedding of
into
is an isometry if
is lower semicontinuous.
Let
and
be matrix order unit spaces, and let
be a unital completely positive linear mapping. Then we have the dual mapping
determined by
. Let
denote the dual space of
.
is just the subspace of
consisting of those
such that
. For any
and
, we have
So
,
and
.
Moreover,
is
-continuous. Let
for
. Then for
,
and
satisfying
, we have
| |
| |
| |
| |
So
is a BW-continuous matrix affine mapping of
into
. In particular,
is a closed matrix convex subset of
.
Clearly
is injective if
is surjective.
Let
be a matrix Lipschitz seminorm on
. On
, we define the matrix gauge
by
Then
for
by Lemma 4.3 in [21] .
Proposition 3.6.
Let
and
be matrix order unit spaces, and let
be a unital completely positive linear mapping which is surjective. Let
be a matrix Lipschitz seminorm on
, and let
be a sequence of the corresponding quotient seminorms on
, defined by
Then
-
(1)
is a matrix Lipschitz seminorm on
;
-
(2)
is a complete isometry for the matrix norms
and
on
and
;
-
(3)
is a complete isometry for the corresponding matrix metrics
and
;
-
(4)
If
is a matrix Lip-norm, then so is
.
-
Proof.
(1) For
, we have
| |
| |
| |
| |
| |
| |
If
,
and
, we have
| |
| |
| |
| |
and
| |
| |
| |
| |
Given
. We have
| |
| |
and so
. If
with
, then
Letting
, where
, we get
Since
and
is positive, we have
Now by Proposition 3.1 in [17] , there exists an
such that
.
Similarly, there exists a
such that
. Therefore,
.
Thus
is a matrix Lipschitz seminorm on
.
(2) Let
. For any
we clearly have
, and so if
we have
Consequently,
. But let
be given, and let
with
. Then there is an
such that
and
. Thus,
. Consequently,
| |
| |
| |
Taking the supremum over
with
, we see that
Since
is arbitrary, we obtain that
. Thus
is a complete isometry.
(3) By (2), we have
| |
| |
| |
where
, that is,
is a complete isometry for the corresponding matrix metrics
and
. (4) Suppose that
is a matrix Lip-norm. Since
is
-continuous,
is surjective, and
is BW-compact,
is a homeomorphism of
onto
. Because
gives the BW-topology on
,
gives the relative topology of
. According to (3),
gives the BW-topology on
. Therefore,
is a matrix Lip-norm. □
Notation 3.7.
Under the conditions of Propositions 3.6 we will say that
induces
via
.
For a matrix convex set
in a locally convex vector space, let
be the set of all matrix affine mappings from
to
(see §6 in [21] ). On the other hand, we have
Proposition 3.8.
Let
be a matrix order unit space, and let
be a compact matrix convex subset of
. View the elements of
as matrix affine mapping from
to
(Proposition 6.1 in [
21]
), and let
consist of their restrictions to
, with
the restriction mapping of
onto
. Then
is a matrix order unit space, and
.
-
Proof.
Clearly, with the natural matrix order structure on
and the order unit
,
is a matrix order unit space.
For
, we define the mapping
by
.
Then
and
for
. Thus
.
Suppose that
and
. By Theorem 1.6 in [19] , there is a
and a self-adjoint
such that
for all
,
, and
So we obtain
for all
and
. Thus
in
. If
for some
, we would then have that
. Thus,
. Therefore,
. □
Notation 3.9.
We will call the matrix order unit space
in the Proposition 3.8 the quotient of
with respect to
, and will identify
with
. When
is a quantized metric space,
is a quantized metric space by Proposition 3.6 .
is called the quotient space of
with respect to
and
.
Proposition 3.10.
Let
,
and
be matrix order unit spaces. Suppose that
and
are unital completely positive linear mappings which are surjective. Denote
. If
is a matrix Lipschitz seminorm on
,
and
are the induced matrix Lipschitz seminorms of
via
and
, respectively, and
is the induced matrix Lipschitz seminorm of
via
, then
.
-
Proof.
This follows by exactly the same argument used for quantum Gromov-Hausdorff distance in [17] . □
4 Quantized Gromov-Hausdorff distance
As in the situation of compact quantum metric spaces, we need a corresponding notion of bridge for estimating distance between quantized metric spaces.
Let
and
be two quantized metric spaces with the matrix norms
and
determined by their matrix orders on
and
, respectively. We form the direct sum
of operator spaces (see §2.6 in [13] ).
becomes a matrix order unit space.
Definition 4.1.
Let
and
be quantized metric spaces.
A matrix bridge between
and
is a matrix seminorm
on
such that
-
(1)
is matrix continuous for the matrix norm
on
, that is, each
is continuous for
on
.
-
(2)
for
and
and
.
-
(3)
but
.
-
(4)
For any
,
and
, there is a
such that
and similarly for
and
interchanged.
Example 4.2.
Suppose
and
are quantized metric spaces. Choose
and
. For
, we define
by
Then
is a matrix seminorm on
, and satisfies the conditions (1), (2) and (3) of Definition 4.1 . For any
and
, choose
. Then we have
and similarly if we are given
. So
is a matrix bridge between
and
.
Proposition 4.3.
Let
be a matrix bridge between quantized metric spaces
and
. Define
on
by
Let
and
be the projections from
onto
and
, respectively, which are unital completely positive linear surjective mappings. Then
is a matrix Lip-norm on
, and it induces
and
via
and
, respectively. If
and
are lower semicontinuous, then so is
.
-
Proof.
For
and
, we have
| |
| |
| |
| |
| |
| |
| |
and
| |
| |
| |
and for
and
, we have
| |
| |
| |
| |
| |
Thus
is a matrix seminorm. Since
for
and
,
for
.
If
, then
, and hence
and
,
, where
. From
and
for
, we have
and so for
,
| |
| |
| |
Thus
. So
is a matrix Lipschitz seminorm.
Similar to the same argument used in Theorem 5.2 of [17] , we have that
induces
and
via
and
, respectively. By Proposition 3.1 in [20] , Proposition 7.5 in [21] and Theorem 5.2 in [17] (see also §2 in [17] ), the
-topology on
agrees with the BW-topology. Therefore,
is a matrix Lip-norm on
.
Suppose that
and
are lower semicontinuous. Clearly,
is lower semicontinuous since
is matrix continuous. □
Notation 4.4.
We will denote by
the set of matrix Lip-norms on
which induce both
and
via
and
, respectively.
By Proposition 4.3 and Example 4.2 ,
. From Proposition 3.6 , we can view
and
as closed matrix convex subsets of
.
Now we introduce our notion of distance for quantized metric spaces.
Definition 4.5.
Let
and
be quantized metric spaces. We define the quantized Gromov-Hausdorff distance
between them by
where
is the Hausdorff distance between
and
for
.
Given a quantized metric space
, we define its diameter
to be the diameter of
with respect to
. The following proposition indicates that the quantized Gromov-Hausdorff distance is always finite.
Proposition 4.6.
Let
and
be quantized metric spaces.
Then
-
Proof.
Choosing arbitrarily
,
, we set
As Example 4.2 ,
is a matrix bridge between
and
. By Proposition 4.3 ,
, where
is a matrix Lip-norm in
. Then for
, and
with
, we have
| |
| |
| |
| |
| |
If
, we get
by Proposition 3.1 in [20] . If
, similar to the proof of Proposition 4.2 in [20] , there are
, such that
Since
, we obtain
| |
| |
| |
| |
Applying the same argument, we have
Hence
The arbitrariness of
implies that
by Proposition 3.1 in [20] . □
It is clear that the quantized Gromov-Hausdorff distance is symmetric in
and
. We come to prove that it satisfies the triangle inequality.
Theorem 4.7.
If
,
and
be quantized metric spaces, then
-
Proof.
Given
. Then there are
and
such that
and
We define
on
by
Then for
and
, we have
| |
| |
| |
| |
| |
and
| |
| |
| |
| |
and for
and
, we have
| |
| |
| |
| |
if and only if
and
, and this is equivalent to that
. Therefore,
is a matrix Lipschitz seminorm.
Similar to the same argument used in Lemma 4.4 of [17] , we have that
induces
,
,
,
and
for the evident quotient mappings by Proposition 3.10 .
By Proposition 3.1 in [20] , Proposition 7.5 in [21] and Lemma 4.4 in [17] (see also §2 in [17] ), the
-topology on
agrees with the BW-topology. So
is a matrix Lip-norm on
.
By Proposition 3.6 , we have
and
So
| |
| |
| |
| |
| |
| |
| |
Since
is arbitrary, we obtain
□
Proposition 4.8.
Let
be a quantized metric space, and let
and
be compact matrix convex subsets of
.
If
is the quotient space of
with respect to
and
,
, then we have
-
Proof.
Let
and
be the projections from
onto the first space
and the second space
, respectively. Denote
and set
,
,
, and
, the BW-closed matrix convex hull of the graded set
. Let
be the quotient of
with respect to
. Then
by Proposition 3.8 .
For
, we have
, that is,
for
.
This is equivalent to
for
,
since
. And this holds if and only if
and
for
and
,
, that is, if and only if
and
.
So
. And thus there is a complete order isomorphism
from
onto
.
Given
. We define a matrix seminorm
on
by
Then
is a matrix bridge between
and
, and
by Proposition 4.3 , where
Thus
is a matrix Lip-norm on
. Let
and
be the quotient space of
with respect to
and
. Then
by Propositon 3.10 .
Since
for
, we have that
. For
, we have
| |
| |
| |
that is,
. Because
, we get that
So
| |
| |
| |
| |
| |
| |
| |
Since
is arbitrary, we obtain the desired inequality. □
Let
and
be compact quantum metric spaces. The quantum Gromov-Hausdorff distance between them is defined by
where the infimum is taken over all Lip-norms
on
which induce
and
(see Definition 4.2 in [17] ).
Proposition 4.9.
Let
for
be compact quantum metric spaces, and let
be an associated quantized metric space of
(see Example 3.5 ). Then
-
Proof.
Suppose
. Then
for
and
for
, where
(see Example 3.5 ). So for
, we have
| |
| |
| |
| |
| |
where
is the projection from
onto
and
is the restriction of
to
. Denote
. Let
be given. Then there is a
such that
. Setting
, we have that
and
| |
| |
| |
| |
Thus
for
. Similarly, we have that
for
. So
.
For
and
, let
and
. Then
and
. Since
, we obtain
| |
| |
| |
(see §2 in [17] ). So
Therefore, by Theorem 4.3 and Proposition 7.1 in [17] , we have
| |
| |
| |
| |
| |
| |
Consequently,
. □
5 Distance zero
In this section, we show that
is equivalent to the existence of a complete isometry between them in the following sense.
If
is a quantized metric space, then
is the largest lower semicontinuous matrix Lip-norm smaller than
by Corollary 4.5 in [21] . From Proposition 7.1 in [21] ,
extends uniquely to a closed matrix Lip-norm
on the subspace
, where
is the completion of
for its matrix norm.
Definition 5.1.
Let
and
be quantized metric spaces. By a complete isometry from
onto
we mean a unital complete order isomorphism
from
onto
such that
, that is,
for all
.
Lemma 5.2.
Let
be a quantized metric space. Then
-
Proof.
Let
be given, and define
for
,
and
. Clearly
is a matrix continuous matrix seminorm on
, and
and
.
For
and
and
, setting
, we have
by Proposition 3.6, Proposition 7.1 and Proposition 3.4 in [21] . Given
and
and
. By Lemma 7.3 in [21] , there is a sequence
of elements in
such that
and
converges to
in norm. Consequently, we can find an
such that
.
So
. Thus
is a matrix bridge between
and
.
Define
for
,
and
. By Proposition 4.3 ,
. For
and
, we have that
, and hence
| |
| |
| |
| |
| |
| |
| |
where
and
are the projections from
onto
and
, respectively. For
and
, there is a
such that
by Arveson's extension theorem. So
| |
| |
| |
| |
| |
| |
| |
Thus
for
, and so
Therefore,
. By the arbitrariness of
, we obtain
By Proposition 3.4 in [21] and the proof of Theorem 4.4 in [21] , we can prove that
similarly. □
Theorem 5.3.
Suppose
and
are quantized metric spaces.
If there exists a complete isometry
from
onto
, then
-
Proof.
For
, we define
for
,
and
. Clearly
is a matrix seminorm on
and
for
,
and
. And we have that
and
. If
and
with
and
, we have that
since
is completely bounded(see Proposition 3.5 in [12] ).
Given
and
. Taking
, we have
| |
| |
While if
and
, we can take
such that
, and hence we have
| |
| |
Therefore,
is a matrix bridge between
and
. Define
for
,
and
. By Proposition 4.3 ,
. For
and
, we have that
, and so
| |
| |
| |
| |
| |
Similarly, for
and
, we have that
. Thus we obtain that
for
, and so
Therefore,
. Since
is arbitrary, we conclude
Now, by Theorem 4.7 and Lemma 5.2 we have
So
. □
Given a quantized metric space
. From Proposition 6.1 in [21] and the proof of Proposition 3.5 in [19] , the mapping
, defined by
for
, is a unital complete order isomorphism from
into
, and
can be extended to a unital complete order isomorphism
from the completion
of
onto
. Define
where
. Then
is a matrix gauge on
.
Denote
Let
and
be the norm closure of
in
. Denote
and
. The matrix gauge
on
determined by
is called the closure of
.
is closed if
on the subspace where
is finite (see Definition 7.2 in [21] ).
Lemma 5.4.
If
is closed, then
for
.
-
Proof.
Denote
and set
. Define
for
, where
is the order unit of
. Here we view
as the subspace of
consisting of those
with
for
. Clearly,
is a matrix gauge on
and
for all
, The generalized bipolar theorem says that
is the smallest weakly closed absolutely matrix convex set containing
(see Proposition 4.1 in [4] ). Since
is a matrix gauge and
is a unital complete order isomorphism,
is absolutely matrix convex. The closeness of
implies that
is normed-closed by Lemma 7.4 in [21] . So
is weakly closed. Thus
For
, we have
| |
| |
| |
| |
| |
| |
and
| |
| |
| |
So
if and only if
for all
and
.
Suppose that
for all
and
. Then for
, we have
| |
| |
| |
| |
| |
Thus
exactly if
for all
and
. According to Lemma 4.1 in [21] ,
exactly if
for all
and
. So
exactly if
| |
| |
| |
| |
for all
and
. Because
,
if and only if
for all
and
. And this says exactly that
. Therefore,
for
. So
for
. □
Lemma 5.5.
Let
and
be quantized metric spaces such that
and
are closed. For every matrix affine mapping
from
onto
which is completely isometric for
and
, there is a unital complete order isomorphism
from
onto
such that
.
-
Proof.
Define
by
for
and
. Since
is isometric and matrix affine,
is well-defined. Clearly,
is unital and surjective. On the level of matrices, we have
for
and
. Since
in
if and only if
for all
and
,
is a unital complete order isomorphism from
onto
. Since
and
are closed,
is a unital complete order isomorphism from
onto
by Lemma 5.4 . That
is completely isometric for
and
implies that
for all
and
. Because
and
are closed, they are lower semicontinuous, so that
on
by Theorem 4.4 in [21] , and similarly for
. Thus
is a unital complete order isomorphism from
onto
such that
. □
Theorem 5.6.
Suppose
and
are quantized metric spaces.
If
then there exists a complete isometry
from
onto
.
-
Proof.
Since
, we have
by Lemma 5.2 and Theorem 4.7 . From that
, there is a sequence
of matrix Lip-norms such that
So for each
, we have
And for
,
, by Proposition 3.6 we have
where
, is the projection from
onto
. Therefore, for each
we get
where
is the Gromov-Hausdorff distance(see Definition 3.4 in [6] ) between
and
. As in the proofs of Theorem 7.6 and Theorem 7.7 in [17] , there is a subsequence
which converges uniformly on the disjoint union
to a semi-metric
and
determines an isometry
from
onto
by the condition that
. Similarly, there is a subsequence
of
which converges uniformly on
to a semi-metric
and
determines an isometry
from
onto
by the condition that
. In general, once
have been choosen, there is a subsequence
of
which converges uniformly on
to a semi-metric
and
determines an isometry
from
onto
by the condition that
.
Given
and
,
, satisfying
, and
. Let
. We can find
such that if
then
for
. Now for
we have
| |
| |
| |
| |
| |
Since
is arbitrary, we have
. But
By Lemma 7.4 in [17] , we obtain
So
is matrix affine.
Now, by Lemma 5.5 we conclude that there exists a unital complete order isomorphism
from
onto
such that
, that is,
is a complete isometry from
onto
. □
6 Completeness
For the metric space of complete isometry classes of quantized metric spaces with the quantized Gromov-Hausdorff distance, we show in this section that it is complete.
Let
be a sequence of matrix order unit space. We will denote by
the operator space direct sum that is formed of all sequences
with
and
, and by
the operator space direct sum of
(see §2.6 in [13] ). Then
and
are matrix order unit spaces.
Suppose we have a sequence
of quantized metric spaces. Suppose further that we have a sequence
of matrix Lip-norms with
.
Define
on
, the full product, by
and set
It is easy to check that
is a self-adjoint subspace of
containing
, and so is a matix order unit space, and that
is a matrix Lipschitz seminorm on
.
For the evident identifications, we have
Given a family of graded sets
. We denote by
the graded set
. If
is a graded set in a vector space, we denote by
the matrix convex hull of
. Let
,
and
.
Proposition 6.1.
and
are BW-dense in
. In particular,
is BW-dense in
for
.
-
Proof.
Since the matrix polar
| |
| |
| |
| |
| |
we have
| |
| |
For
and
, we have that
, and so
. Thus
that is,
. If
and
, then
and so
. Thus
. Clearly,
. Therefore,
By the bipolar theorem in matrix convexity (see Corollary 5.5 in [5] ),
is BW-dense in
. Evaluting the completely positive mappings at
, we see that
is BW-dense in
. Because
,
is BW-dense in
. □
Define
on
by
for
. Similar to the proof of Proposition 4.3 , we have
Proposition 6.2.
is a matrix Lip-norm on
, and induces
and
and
,
, via the evident projections.
For
and
, set
. Since
induces
, we can find
such that
and
, where
is the evident projection from
onto
. Similarly, we can find
such that
and
. Continuing in this way, for
we get
such that
and
. We let
be the unique element of
such that
for
. Then
. So,
. Set
.
Since
induces
, via the evidence projections,
For
with
and
, we can find
with
Similarly, we can find
with
Inductively, we can find
with
and
for
. Consequently,
Similarly, for
we can find a
such that the inequality above holds. Thus by Proposition 6.1, we have
Proposition 6.3.
For
, we have
-
Proof.
For
, we can find a
such that
from the discussion before the proposition. Suppose
and
. For each
,
is a BW-closed subset of
, and
for all isometries
. From Proposition 6.1 , the BW-closure
of
is
, and so by Theorem 4.6 and Theorem 4.3 in [19] , there exist
and
for
satisfying
such that
For each
, we can find
so that
Thus
| |
| |
| |
| |
| |
| |
because
is a matrix metric (see Example 5.2 in [21] ). Since
is arbitrary, we obtain the desired inequality. □
Now for
, there are
with
and
So
| |
| |
where
is the diameter of
with respect to
.
By Proposition 5.2 and Proposition 3.8 in [20] , we have
where
. So there is an
such that
Set
Then
is a non-empty closed bounded subset of
. Clearly,
.
So there exists a
. We have
Thus
, and we obtain
Proposition 6.4.
For
,
induces
via the evident projection.
Theorem 6.5.
The metric space
of complete isometry classes of quantized metric spaces, with the metric
, is complete.
-
Proof.
Let
be a sequence in
which is Cauchy with respect to the quantized Gromov-Hausdorff distance
. To show that
converges it suffices to show that a subsequence converges. Since
is Cauchy, we can choose a subsequence, still denoted by
, such that
for all
. By definition, there exist
with
for all
. It follows that
Let
be given. Then there is an
such that
By Proposition 6.2 , Proposition 6.3 and Proposition 6.4 , we have
| |
| |
| |
for
. This says that
is
-dense for
in
. But
is BW-compact for the topology from
by Proposition 6.2 . Thus
is totally bounded for
, and so
is totally bounded for
. Let
be the completion of
for
. We let
denote also the extension of
to
. Then
is a compact matrix convex set. For
, we have
and
| |
| |
| |
where
,
, and
, and
satisfying
and
. So the map
, given by
for
,
and
satisfying
, is well-defined and
can be extended to an element
. Moreover if
in
then
in
and
. Thus
can be regarded as a matrix order unit subspace of
.
Define the map
, by
for
and
. Clearly,
is continuous. For
with
,
satisfying
, we have
that is,
. Since
is dense in
and
is compact, we obtain that
.
If
with
and
, we can find
such that
. Thus
. So we can find
with
and
. But
so that
. Thus
. Therefore,
is injective. So
is a homeomorphism of
onto
for
. From this we see that the
-topology on
agrees with the BW-topology. Hence
is a matrix Lip-norm on
.
By Proposition 6.2 and Proposition 6.4 , we obtain
which indicate that, for
,
is a Cauchy sequence for
, and has a limit
. Clearly
is a compact matrix convex set.
Because
is completely order isomorphic to a dense subspce of
(Proposition 6.1(1) in [21] ), we can view
as a dence subspace of
. Let
be the map which restricts the elements of
to
and
. Then
is a quantized metric space.
Given
. Then there is an
such that
For
, we have
| |
| |
| |
for
. Letting
, we obtain
for
, and so
. By Proposition 4.8 , for
we have
Therefore,
. □
7 Matrix approximability
In this section, we establish a matrix approximability theorem for
-exact matrix order unit spaces.
Lemma 7.1.
Let
be a quantized metric space and let
,
with
for
.
Suppose
for
. Then
.
-
Proof.
By Proposition 2.11 in [8] , we have
| |
| |
| |
| |
| |
□
An operator space
is said to be
-exact if for every finite-dimensional subspace
and
there is an isomorphism
from
onto a subspace of a matrix algebra such that
. A matrix order unit space
is said to be
-exact if it is
-exact as an operator space.
Theorem 7.2.
Let
be a quantized metric space. If
is
-exact, then for every
, there is a quantized metric space
such that
-
Proof.
Since
is
-exact, by Lemma 5.1 in [9] there is a unital complete order embedding
and a net
of unital completely positive mappings through matrix algebras such that
converges pointwise to
. Given
. By Lemma 7.2, we have
where
. From Proposition 7.5 in [21] ,
is totally bounded for
. So there is a
such that
Denote
and
for
and
.
We define
It is clear that
is a matrix seminorm on
and satisfies the conditions (1), (2) and (3) of Definition 4.1 . For
and
, we can choose
. Then
For
and
, we can take
such that
and
. Then
So
is a matrix bridge between
and
. Define
Then
by Proposition 4.3 . If
, we have
and
| |
| |
| |
On the other hand, if
,
can be extended to a
by Arveson's extension theorem. We have
and
| |
| |
| |
| |
| |
| |
| |
| |
| |
So we obtain that
.
Since
is finite-dimensional,
is a normed-closed (and hence weakly closed) absolutely convex set in
, and
is a normed-closed (and hence weakly closed) absolutely matrix convex set in
. Then for the corresponding matrix seminorm
of the maximal envelope
of
in
(see Example 3.4 ), we have
(see page 181 in [4] ). It is clear that
is a matrix Lipschitz seminorm. Since the image of
in
is totally bounded for
and
, the image of
in
is totally bounded for
. By Theorem 5.3 in [20] ,
-topology on
agrees with the BW-topology. So
is a matrix Lip-norm on
. By Lemma 3.2.3 in [1] , there is a (real linear) projection
from
onto
with
. We define
by
for
. Then
is a bounded linear mapping with
. Define
It is clear that
is a matrix Lip-norm on
since
for all
and
is a matrix Lip-norm. And for
, we have
| |
| |
Define
It is clear that
is a matrix seminorm on
and satisfies the conditions (1), (2) and (3) of Definition 4.1 . For
and
, we choose
. Then we have
For
and
, we choose
. Then we have
So
is a matrix bridge between
and
. Define
Then
by Proposition 4.3 . For
,
and
| |
| |
| |
| |
For
,
can be extended to a
by Arveson's extension theorem. We have
| |
| |
| |
| |
So
. Therefore,
□
8 Sphere as the limit of matrix algebras
Let
be a connected compact semisimple Lie group with a continuous length function
on
, which satisfies the additional condition
for all
. Fix an irreducible unitary representation
of
. Then
have a highest weight vector
, of norm
, unique up to a scalar multiple. Let
be the rank-one projection for
. Denote by
the stability subgroup for
under
. For any
, we form the
inner tensor power
of
.
Let
denote the subrepresentation generated by
. Then
is irreducible with
as highest weight vector. We let
.
The action of
on
by conjugation by
is denoted by
. We let
denote the action of
on
, and so on
, by left-translation. We denote the corresponding Lip-norm for
and
on
by
, that is,
and we denote the Lip-norm for
and
on
by
, that is,
here we view
as a subalgebra of
. By Theorem 3.2 in [18] , the quantum metric spaces
converge to
for quantum Gromov-Hausdorff distance as
goes to
. In this section, a more general statement is established.
Let
be the matrix norm on
. Set
, where
and
, where
Then
and
are quantized metric spaces for all
by Example 6.5 in [21] . As in [18] , we will not restrict
to the Lipschitz functions. Let
denote the rank-one projection for
. We denote the corresponding Berezin symbol mapping from
to
by
. Then
is unital, positive, norm-nonincreasing and
-
-equivariant (see page 73 in [18] ). For
and
, define
For
, define
and denote
.
Lemma 8.1.
For any
, we have
-
Proof.
Since
is a unital positive mapping from
to
,
is unital completely positive and hence
by Theorem 3.8 and Proposition 3.5 in [12] . So we have
| |
| |
| |
by the
-
-equivariation of
. □
Put on
the inner product from
, while on
its Hilbert-Schmidt inner product. Then the mapping
from
to
has an adjoint operator
from
to
. For any
, a function
such that
is called a Berezin contravariant symbol for
. Moreover,
is unital, positive, norm-nonincreasing, and
-
-equivariant (see page 75 in [18] ). From Theorem 3.10 and Proposition 3.5 in [12] ,
is unital completely positive and
.
So by the same argument as in the proof of Lemma 8.1 , we obtain:
Lemma 8.2.
For any
, we have
where
.
Denote
for
,
, and
where
denotes the usual (un-normalized) trace on
and
is the dimension of
. Set
where every
is naturally identified with an element
of
. Then:
Lemma 8.3.
For
, we have
-
Proof.
Suppose
. Then for any
, we have
| |
| |
| |
| |
| |
| |
by the formula (2.2) in [18] . So
□
Since the sequence
converges to
as
(see page 80 in [18] ), there is an
such that
for
. So we obtain:
Proposition 8.4.
For
,
is a matrix bridge between
and
, and hence
, where
From Theorem 6.1 in [18] , we have:
Lemma 8.5.
There is an
such that
for all
and
.
Theorem 8.6.
With notation as above, the quantized metric spaces
converge to
for quantized Gromov-Hausdorff distance as
goes to
.
-
Proof.
Given
. Choose
. Then for
, we have that
by Proposition 8.4 . Given
. we have
, and
| |
| |
| |
| |
| |
On the other hand, if
, then
, and
| |
| |
| |
| |
| |
| |
| |
| |
| |
by Lemma 8.5 . Therefore, for
, we have
that is,
. □
Acknowledgements
I am grateful to Marc Rieffel for valuable discussions. I also thank Hanfeng Li for helpful comments. This research was partially supported by Shanghai Priority Academic Discipline, China Scholarship Council, and National Natural Science Foundation of China. References
-
B. Blackadar, E. Kirchberg, Generalized inductive limits of finite-dimensional
-algebras, Math.Ann. 307 (1997), 343–380.
-
M. -D. Choi, E. G. Effros, Injectivity and operator spaces, J. Funct. Anal., 24 (1977), 156–209.
-
A. Connes, Compact metric spaces, Fredholm modules, and hyperfiniteness, Ergodic Theory Dynam. Systems,9(2) (1989), 207–220.
-
E. G. Effros and C. Webster, Operator analogues of locally convex spaces, Operator algebras and applications(Samos, 1996), 163–207, NATO Adv. Sci. Inst. Ser. C Math. Phys. Sci., 495, Kluwer Acad. Publ., Dordrecht, 1997.
-
E. G. Effros, S. Winkler, Matrix convexity: operator analogues of the Bipolar and Hahn-Banach theorems, J. Funct. Anal., 144 (1997), 117–152.
-
M. Gromov, Metric structures for Riemannian and non-Riemannian spaces, Birkhäuser Boston Inc., Boston, MA, 1999.
-
R. V. Kadison, A representation theory for commutative topological algebra. Mem. Amer. Math. Soc., 7, 1951.
-
D. Kerr, Matricial quantum Gromov-Hausdorff distance. J. Funct. Anal., 205 (2003), 132–167.
-
D. Kerr, H. Li, On Gromov-Hausdorff convergence for operator metric spaces, arXiv:math.OA/0411157 v2, 2004.
-
H. Li,
-deformations as compact quantum metric spaces, to appear in Comm. Math. Phys., arXiv:math.OA/0311500.
-
N. Ozawa, M. A. Rieffel, Hyperbolic group
-algebras and free-product
-algebras as compact quantum metric spaces, to appear in Canad. J. Math., arXiv:math.OA/0302310.
-
V. I. Paulsen, Completely bounded maps and dilations, Pitman Research Notes in Mathematics Series, 146. Longman Scientific & Technical, Harlow; John Wiley & Sons, Inc., New York, 1986.
-
G. Pisier, Introduction to operator space theory, London Mathematical Society Lecture Note Series, 294. Cambridge University Press, Cambridge, 2003.
-
M. A. Rieffel, Metrics on states from actions of compact groups, Doc. Math., 3 (1998), 215–229.
-
M. A. Rieffel, Metrics on state spaces, Doc. Math., 4 (1999), 559–600.
-
M. A. Rieffel, Group
-algebras as compact quantum metric spaces, Doc. Math., 7 (2002), 605–651.
-
M. A. Rieffel, Gromov-Hausdorff distance for quantum metric spaces, Mem. Amer. Math. Soc. 168 (2004), 1–65.
-
M. A. Rieffel, Matrix algebras converge to the sphere for quantum Gromov-Hausdorff distance, Mem. Amer. Math. Soc. 168 (2004), 67–91.
-
C. Webster, S. Winkler, The Krein-Milman theorem in operator convexity, Trans. Amer. Math. Soc., 351 (1999), 307–322.
-
W. Wu, Non-commutative metric topology on matrix state spaces, to appear in Proc. Amer. Math. Soc., arXiv:math.OA/0410587.
-
W. Wu, Non-commutative metrics on matrix state spaces, arXiv:math.OA/0411475.
Department of Mathematics, East China Normal University, Shanghai 200062, P.R. China E-mail address : wwu@math.ecnu.edu.cn Current address : Department of Mathematics, University of California, Berkeley, CA 94720-3840 E-mail address : wwu@math.berkeley.edu