Continuous Bijection Not Homeomorphism Fixed Point
1. Introduction
A paper [1] (resp. [2]) explored some features of
(resp. Khalimsky line) and newly proposed two topologies derived from quotient spaces of the one point compactification of
(resp. Khalimsky line). Based on the work, the present paper continues the study of
from the viewpoint of fixed point theory and finally addresses a certain unsolved problem already mentioned in abstract.
From now on the term "Marcus-Wyse" will be often replaced with "
-" for brevity. We shall also use the symbol "
" to define new terminology. In a category
, let us recall that an object
has the fixed point property (FPP) if for every self-morphism
of X, there is an element
such that
. After establishing a certain category relating to the Alexandroff compactification of the M-topological plane, we address the unsolved problem: Under what category does
have the fixed point property (FPP, for short) ? Regarding this work, we need some new ideas and tools.
Let us now recall some terminology which will be used later in the paper. We usually denote by
the M-topological plane (see Section 2), where
is the topological structure of
generated by a special kind of set as a base (see (1) and (2) for the details). As usual, we often denote by
the first infinite cardinal number. Besides,
means the cardinality of a given set, and we denote by
the set of natural numbers. As usual, a topological space
is called locally compact [3] if for each point
there is a compact neighborhood containing the point x. It is obvious that the M-topological space
is neither a Hausdorff nor a compact space but it is a locally compact space (see later in Section 2 for the details). Hence we can establish the Alexandroff one point compactification of
and further, we call it the infinite M-topological sphere [3]. Hereafter, we denote by
the one point compactification of
[1].
To study the FPP of
, we need to define the following category.
Definition1.
We denote by the category whose object is the only and morphisms are all continuous self-maps g of such that with or is a singleton.
The present paper disproves the FPP of the infinite M-topological sphere in
which denotes the set of all continuous self-maps of the infinite M-topological sphere (see Theorem 2). Hence we need to establish a suitable subcategory of it having the FPP. Although, in general, a continuous bijection need not imply a homeomorphism, we wonder if a continuous self-bijection of
is a homeomorphism. Thus we may raise the following queries.
(1-1) Does any continuous self-map g of
have a point x in
such that
?
(1-2) Is a continuous self-bijection of
equivalent to a self-homeomorphism of it?
(1-3) Under what category does
have the FPP?
(1-4) Does
have the FPP in the category
?
(1-5) What category is the maximal one with respect to the inclusion relation which admits the FPP of
?
The present paper suggests some affirmative solutions to the queries (1-2), (1-3), and (1-4) and answers to the query of (1-1) negatively. In view of these facts, comparing with the non-FPP of the Hausdorff compactifications of the 2-dimensional usual topological space, we can recognize the study of the FPP of
has its own feature.
The remainder of the paper is organized as follows: Section 2 deals with some notions related to the M-topology and the Alexandroff compactification. Section 3 investigates various properties of continuous self-maps (or surjections or bijections) of
. Section 4 proves that every morphism g of
has some point
such that
. Section 5 concludes the paper with some remarks and further work.
2. Preliminaries
A topological space
is called an Alexandroff space if each point
has the smallest open set in
[4]. Indeed, it is obvious that the usual topological space
is not an Alexandroff space. As an Alexandroff topological space [4,5], the M-topological space was proposed [6] and the study of various properties of it includes the papers [1,6,7,8,9,10,11,12,13]. Regarding digital spaces [14] in
, we will follow the concept of a digital k-neighborhood of a point
.
For a point
we follow the notation [15].
Then we use the set
developing the M-topology on
, where
The M-topology "
" on
is derived from the set
in (1) as a base [6]. Then we use the notation
for the topological space. To further state a point in
, we call a point
double even if
is an even number such that each
is even,
; even if
is an even number such that each
is odd,
; and odd if
is an odd number [6].
In a subspace of
, an odd point (resp. a double even point or even point) is denoted by a black large dot (resp. the symbol ◊). According to (1), under
, it appears that the singleton being composed of "
" is a closed set and the singleton consisting of a black large dots is an open set. Besides, we will denote by
(resp.
) the set of all odd points (resp. double even or even points) in
. In addition, for a set
, we denote by
, called an M-topological space, the subspace induced by
. Owing to (1), it is obvious that
is an Alexandroff space.
Under
, the smallest (open) neighborhood of the point
of
, denoted by
, is determined according to the given point p, as follows:
From now on, for a point p in
, we follow the notation
or
[7]. Owing to (2),
is obviously a locally finite topological space.
According to the property (2), the following properties are obtained in
:
Lemma1.
(1) A subset D of is open [1] if and only if
(2) A subset C of is closed if and only if whenever , where
By Lemma 1 and the property (2), under
, we obtain the following: For the point
,
can be represented as follows:
where "
" means the closure operator.
Owing to (1),
is locally compact but neither compact nor Hausdorff. To be specific, let us check that
is not Hausdorff. To be precise, under the M-topological space
, take two distinct points
. Then we see that no open sets
exist such that
,
and
because the smallest open set
is
and the smallest open set
is the singleton
(see Section 2 for the details), i.e.,
, which implies that
is not Hausdorff.
Next, we now check the non-compactness of
. Let
Though C is an open covering for
with
, a deletion of at least one element from C does not support a covering of
, which implies that
is not compact. Finally, since any point
has an smallest open set which is compact, it turns out that
is locally compact. Thus we can proceed with the Alexandroff compactification of
[3], denoted by
.
To be specific, take a certain object outside X, denoted by the symbol
, and add it to X, creating the set
[4]. Topologize
using the following sets as open sets '
' in
.
, where
is a closed and compact subset of
, and
means the complement of the set O in
.
This topological space is called the Alexandroff one-point compactification of
and denoted by
.
Taking this approach, we obtain
from
which will be often called the infinite M-topological sphere [1]. In
, it appears that
Besides, under
, we need to remind that for an open set, say O, containing the point ∗,
is a closed and compact set of
. Owing to the property (2), it appears that
and further, for an open set
,
should be finite and closed in
.
Proposition1.
Under , we obtain the following:
(1) is closed but is not compact, where A is a countable subset of .
(2) is not open, where B is a denumerable subset of .
Proof.
(1) It is clear that
is closed, where
. Obviously, the set
is an intersection of the closed sets
and M is a countable set. Thus the proof is completed.
Owing to (4), the set
is not compact in
because it is denumerable.
(2) Let us consider the set
in
. Whereas it is closed in
, it is not compact in
, which proves the assertion. □
In view of Proposition 1(2), we obtain the following:
Corollary1.
is not open.
Proof.
Though
is closed in
, it is not compact in
, which completes the proof. □
Remark1.
is connected.
Proof.
According to the definition of
, we obtain
. Owing to both the connectedness of
and this property lead to the connectedness of
. □
For two spaces
and
, a map
is said to be M-continuous at a point
if g is continuous at the point x from the viewpoint of M-topology. Furthermore, we say that a map
is M-continuous if it is M-continuous at every point
. Indeed, since
is an Alexandroff space (see the property (1)), we can represent the M-continuity of g at a point
[7], as follows:
In addition, for two spaces
and
, a map
is called an M-homeomorphism if h is an M-continuous bijection and further,
is M-continuous.
3. Continuous Self-Maps of
Let us now consider continuous self-bijections of
which can play important roles in Section 4 (see Theorem 2).
Remark2.
The followings are continuous self-bijections of
.
Proof.
(1) Given the self-bijection
of
, for any point
, we obviously have the following cases.
(Case 1) For any point
and the smallest open set
in
, we have
in
such that
. Then this open set
supports the continuity of
at the point p (see the property (1)).
(Case 2) For the point
and any open neighborhood of
, denoted by
, we have an open neighborhood of ∗, denoted by
in
, such that
because we may take the open sets
and
such that
At the moment, we need to mention that the open set
(resp.
) is not a minimal open set of the corresponding point
(resp. ∗) because both the points ∗ and
do not have their smallest open neighborhoods in
(for the details, see the proof of Lemma 3 and Corollary 4 in the present paper).
(2) Using a method similar to the proof of (1), we complete the proof. □
Lemma2.
A continuous self-map g of such that is not injective.
Proof.
Assume a continuous self-map g of
such that
. Take the smallest open neighborhood of the point p, i.e.,
. Since the set
it has an infinite cardinality because
which implies that the set
of (7) should be finite (see the property (4)). Hence we should have
. Thus the map g should not be injective because
is finite for any point
. □
Let us now investigate a relation between a continuous self-bijection of
and a self-homeomorphism of
.
Theorem1.
A continuous self-bijection of is a homeomorphism.
Proof.
Owing to Lemma 2, every continuous self-bijection g of
has the property
where
,
.
Next, using a method similar to the approach of Remark 2 (Case 1), for the point
and each open neighborhood of
, denoted by
, we have an open neighborhood of ∗, denoted by
, such that
At the moment, we need to mention that the open neighborhood of
(resp. ∗) such as
(resp.
) is not minimal open sets of the corresponding points because the points ∗ and
do not have their minimal open neighborhoods in
. Thus the inverse of the given continuous self-bijection g of
also has the property
where
. Thus, the given map g should be a homeomorphism. □
In view of Lemma 2 and Theorem 1, we obtain the following:
Corollary2.
A self-homeomorphism g of has the property .
As a result, if we consider the category,
, whose object is the only
and morphisms are all self-homeomorphisms of it. Then
has the FPP in
.
4. The Fixed Point Property of in the Category
In view of Theorem 1 and Corollary 2, we may wonder if every continuous self-map g of the space
has some point
such that
. A recent paper [1] proved that
has the FPP in the category of continuous self-bijections of the space. In view of this result, this section proves that
does not have the FPP (see Theorem 1 below), which answers to the query (1-1). It also addresses the queries (1-3) and (1-4) previously posed in Section 1. Let
be the category whose object is the singleton consisting of
and morphisms is the set of all continuous self-maps of
. Namely, after firstly proving the non-FPP of
, we secondly find a certain subcategory of the category of
which admits the FPP in the category. This section mainly focuses on establishing a category in which the FPP of
holds. Eventually, we prove that every morphism g of
in the category
has a point
such that
(see Theorem 3). This is a positive answer to the question in [1] such as "Under what condition does
have the FPP ?". Let us now address the query (1-1) in Section 1.
Theorem2.
does not have the FPP in .
Proof.
Using a counterexample, we prove this assertion. With
, consider the self-map g of
satisfying the following properties of (1-1) and (1-2):
(1-1)
, where
and further,
(1-2) for
with
(see the map g of Figure 1)
Then the map g is continuous because for the closed singleton
is closed in
and further, the inverse image of any smallest open subset
by g is also open in
. Then we observe that the map g does not have a point
such that
. □
According to Theorem 2, we have the following:
Corollary3.
Consider the continuous self-map g of such that
Then not every map g of has a point such that .
To address the queries (1-3) and (1-4) in Section 1, we need the following lemmas:
Lemma3.
([1]) In , the point ∗ does not have an open set that is homeomorphic to , where .
Owing to Lemma 3, we have the following:
Corollary4.
is not an Alexandroff space.
Owing to Remark 1 and the definition of
, we have the following:
Lemma4.
There is no continuous self-map g of such that
Proof.
Suppose a continuous self-map g of the
satisfying the property (10). Then we have the following properties of g:
To be precise, owing to Remark 1 and the continuity of g, we obtain the property (a). Owing to the hypothesis of (10), the complement of
in
is denumerable. Thus
is not open in
. Hence we have the property (b) because
is not compact in
. Using a method similar to the proof of (b), we obviously have the property (c). Finally, according to the property (11), we conclude that
should be an empty set, which invokes a contradiction of being a map of g. □
According to Lemma 4, it turns out that the map g satisfying the hypothesis of (10) is not a continuous self-map of
. Besides, the case
was already referred to in (9) by saying that any continuous self-map satisfying the property (12) does not support the
of
. Hence we may consider the following:
Lemma5.
There is no continuous self-map g of such that
Proof.
Suppose a continuous self-map g of
satisfying the property (13). Namely, assume a point
such that
. For convenience, put
. Then there is an open set
(see the just above of Lemma 1) such that
Meanwhile, we have
. Thus the map g should map the finite set
onto the infinite set
, which invokes a contradiction. □
In view of Corollary 3, and Lemmas 4 and 5, and the property (12), we need to prove the following:
Theorem3.
has the FPP in .
Proof.
(Case 1) With the hypothesis, in case g is a constant map, the proof is completed.
(Case 2) Assume an arbitrary morphism g of
which is not a constant map. Then we prove that there is a point
such that
. For each morphism
of
, we need to check the following cases.
(Case 2-1) In case
, which completes the proof.
(Case 2-2) Let us now suppose the morphism g having the property
. Namely, assume the case
so that we may consider the following two cases.
(Case 2-2-1) Assume the case
. Since the set
is compact and closed in
, it is also closed in
. Hereafter, for convenience, put
. Then, by using the method suggested in Lemma 5, it appears that the point
has
as a subset of
(see the property (1)). Owing to the given morphism g, the set
is a denumerable open set in
and further,
should be a finite set in
. Thus the remaining finite set
should be mapped by the map g onto
. Since the set
is infinite, we have a contradiction (see the map g). Namely, the existence of
such that
invokes a contradiction.
(Case 2-2-2) Let us assume the case
with
. Whereas the set
is open in
, the set
is open in
containing the point ∗. By using a method similar to the proof of (Case 2-2-1) or Lemma 5 above, we have a contradiction to being a map g.
Based on these cases, we can wrap up that every morphism g of
in
has the property
, which leads the FPP of
in
. □
Example1.
Let us consider any map . Then, by Lemma 5 and Theorem 3, there is at least the point such that , which implies that in , has the FPP.
Remark3.
In view of Theorem 2 (in particular, the property (9)), Corollary 3, and Lemmas 4 and 5, the conditions
of Theorem 3 cannot be omitted.
A paper [11] defined the notion of
-homotopy (see Definition 11 of [11]) which can be used in applied topology. The definition is correct. In Figure 2a,b of [11] related to Example 4.1(1) of [11], since there was misprinted, let us now make it correct, as follows (see Figure 2 in the present paper):
Example2.
(1) Example 4.1(1) of [11] is written as follows:
Let us consider the space in Figure 2a,b. Then each of the two processes presented by the dotted arrows on the space as in Figure 2(1)a,b represents an -homotopy of X relative to .
However, there are misprinted figures (a) and (b) for X as shown in Figure 2(1)a,b in the present paper. The corrected one is the following as stated in Figure 2(2)a in the present paper. Using the arrows in Figure 2(2)a in the present paper, we can see an -homotopy of X relative to .
(2) The picture in Figure 2 of [8] is also misprinted with respect to the dotted arrows. The authors correct it with Figure 2(4) instead of Figure 2(3) (check only the dotted arrows).
In addition, in the configuration of Figure 4c of [16], the word "A K-retractible space" should be written by "A non-K-retractible".
5. Concluding Remarks
After finding the category
, we have addressed the unsolved problem: Under what category does
have the FPP? Owing to this approach, we conclude that
is the maximal category which allows for the FPP of the
(see the query (1-5)). Since
is one of the digital models for the typical sphere in traditional geometry, this study of
facilitate the studies of pure and applied mathematics.
As a further work, based on the tools used in this paper, we can expand the obtained results in this paper to the study of newly-established topology such as the topology on
introduce in the paper [17]. Besides, we can further study the FPP problem for a non-compact, non-Hausdorff, and locally compact Alexandroff plane. In addition, we can adapt this approach into the study of combinatorial topological structures involving digital images [18,19,20]. Motivated by a certain cone metric on a digital image [21,22,23], we can establish a metric space derived from an M-topological space.
Author Contributions
Conceptualization, S.-E.H.; methodology, S.Ö.; validation, S.-E.H. and S.Ö.; formal analysis, S.-E.H.; writing—original draft preparation, S.-E.H.; writing—review and editing, S.-E.H. and S.Ö.; visualization, S.Ö.; supervision, S.-E.H.; project administration, S.-E.H.; funding acquisition, S.-E.H. All authors have read and agreed to the published version of the manuscript.
Funding
The first author was supported by Basic Science Research Program through the National Research Foundation of Korea(NRF) funded by the Ministry of Education, Science and Technology (2019R1I1A3A03059103). In addition, this research was supported by "Research Base Construction Fund Support Program funded by Jeonbuk National University in 2020".
Conflicts of Interest
The authors declare no conflict of interest.
References
- Han, S.-E. Topologies of the quotient spaces induced by the M-topological plane and the infinite M-topological sphere. Topol. Appl. 2019, 264, 201–209. [Google Scholar] [CrossRef]
- Han, S.-E.; Na, I.-K. Topologies associated with the one point compactifications of Khalimksy topological spaces. Topol. Appl. 2018, 241, 333–344. [Google Scholar] [CrossRef]
- Munkres, J.R. Topology a First Course; Prentice-Hall, Inc.: Upper Saddle River, NJ, USA, 1975. [Google Scholar]
- Alexandorff, P. Diskrete Räume. Mat. Sb. 1937, 2, 501–518. [Google Scholar]
- Alexandorff, P.S. Uber die Metrisation der im Kleinen kompakten topologischen Räume. Math. Ann. 1924, 92, 294–301. [Google Scholar] [CrossRef]
- Marcus, D.; Wyse, F. Solution to problem 5712. Am. Math. Mon. 1970, 77, 1119. [Google Scholar] [CrossRef]
- Han, S.-E. Generalizations of continuity of maps and homeomorphisms for studying 2D digital topological spaces and their applications. Topol. Appl. 2015, 196, 468–482. [Google Scholar] [CrossRef]
- Han, S.-E. Almost fixed point property for digital spaces associated with Marcus-Wyse topological spaces. J. Nonlinear Sci. Appl. 2017, 10, 34–47. [Google Scholar] [CrossRef][Green Version]
- Han, S.-E. Fixed point property for digital spaces. J. Nonlinear Sci. Appl. 2017, 10, 2510–2523. [Google Scholar] [CrossRef][Green Version]
- Han, S.-E. Homotopic properties of an M A-digitization of 2D Euclidean spaces. J. Comput. Syst. Sci. 2018, 95, 165–176. [Google Scholar] [CrossRef]
- Han, S.-E.; Yao, W. Homotopy based on Marcus-Wyse topology and its applications. Topol. Appl. 2016, 201, 358–371. [Google Scholar] [CrossRef]
- Han, S.-E.; Yao, W. An MA-digitization of Hausdorff spaces by using a connectedness graph of the Marcus-Wyse topology. Discret. Appl. Math. 2017, 216, 335–347. [Google Scholar] [CrossRef]
- Šlapal, J. Digital Jordan curves. Topol. Appl. 2006, 153, 3255–3264. [Google Scholar]
- Herman, G.T. Oriented surfaces in digital spaces. CVGIP 1993, 55, 381–396. [Google Scholar] [CrossRef]
- Rosenfeld, A. Digital topology. Am. Math. Mon. 1979, 86, 76–87. [Google Scholar] [CrossRef]
- Han, S.-E. The fixed point property of non-retractible topological spaces. Mathematics 2019, 7, 879. [Google Scholar] [CrossRef][Green Version]
- Han, S.-E.; Jafari, S.; Kang, J.M. Topologies on which are not homeomorphic to the n-dimensional Khalimsky topological space. Mathematics 2019, 7, 1072. [Google Scholar] [CrossRef][Green Version]
- Kovalevsky, V.A. Finite topology as applied to image analysis. Comput. Vis. Graph. Image Process. 1989, 46, 141–161. [Google Scholar] [CrossRef][Green Version]
- Latecki, L. Topological connectedness and 8-connectedness in digital pictures. Comput. Vis. Graph. Image Process. 1993, 57, 261–262. [Google Scholar] [CrossRef]
- Ptak, P.; Kropatsch, V.G. Nearness in digital images and proximity spaces. DGCI:LNCS 2000, 1953, 69–77. [Google Scholar]
- Aleksić, S.; Kadelburg, Z.; Mitrović, Z.D.; Radenović, S. A new survey: Cone metric spaces. J. Int. Math. Virtual Inst. 2019, 9, 93–121. [Google Scholar]
- Klen, R.; Manajlović, V.; Simić, S.; Vuorinen, M. Bernoulli inequality and hypergeometric functions. Proc. Am. Math. Soc. 2014, 142, 550–573. [Google Scholar] [CrossRef]
- Todorć, V. Harmonic Quasiconformal Mapings and Hyperbolic Type Metrics; Springer: Cham, Switzerland, 2019. [Google Scholar]
Figure 1. Configuration of a continuous self-map g of
such that
with the property (9),
.
Figure 2. Some corrections of the picture of Figure 4 of [11] and the picture of Figure 2 of [8].
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Source: https://www.mdpi.com/2227-7390/8/4/599/htm
0 Response to "Continuous Bijection Not Homeomorphism Fixed Point"
Post a Comment