Review Article Volume 7 Issue 2
Department of Mathematics, College of Natural and Applied Sciences, Mbeya University of Science and Technology, Tanzania
Correspondence: Lucas Wangwe, Department of Mathematics, College of Natural and Applied Sciences, Mbeya University of Science and Technology, Tanzania
Received: April 12, 2023 | Published: May 22, 2023
Citation: Wangwe L. Fixed point theorems for relation-theoretic F-interpolative in branciari distance with an application. Phys Astron Int J. 2023;7(2):122-129. DOI: 10.15406/paij.2023.07.00296
This paper proves fixed point theorems for relation-theoretic F -interpolative mapping endowed with binary relation in Branciari Distance. Henceforth, the results obtained will be verified with the help of illustrative examples. Also, we demonstrate the results with an application in matrix equations.
Mathematics Subject Classification 2010: 47H10, 54H25.
Keywords: Fixed point, interpolative mapping, F -contraction, metric space, binary relation, matrix equation.
Fixed point theory is a fascinating area of research for the researchers studying non-linear phenomena. It has many applications for non-linear functional analysis, Approximation theory, Optimization Theory (Saddle function), Variation inequalities, Game theory (Nash equilibrium) and Economics (Black Scholes theorem). Fixed point theory is quite and sequel to the existing theory of Differential, Integral, Partial, Fractional differential, functional equations and matrix equations. Fixed point theory as well as Banach contraction principle have been studied and generalized in different spaces and various fixed point theorems are developed. In 1968 Kannan1 introduced a discontinuity of contraction mappings that can possess a fixed point on a complete metric space by filling the gap created by Banach for more than thirty years. Reich proved the fixed point theorem using three metric points by combining the concept of Banach and Kanann on complete metric space. Dass-Gupta proved the results of the fixed point theorem of the rational type operator by using contraction mapping in metric space.
In 2000, Branciari2 introduced a class of generalized metric spaces by replacing triangular inequality with similar ones which involve four or more points instead of three and improved Banach contraction mapping principle. In 2008, Azam and Arshad3 using the concept of Branciari investigated the mappings given by Kannan by applying the rectangular properties in a generalized metric space. In 2011, Moradi and Alimohammadi4 generalized Kannan’s results, by using the sequentially convergent mappings and rectangular properties in metric space. Furthermore, Morandi and Alimohammadi investigated and extended Kannan’s mapping by using the ideal due to Branciari. Since then, several authors involved in investigations of Banach’s contraction mappings using rectangular properties in different spaces.
In 2004, Ran and Reurings5 proved an order-theoretic analogue of Banach contraction principle which marks the beginning of a vigorous research activity. This result was discovered while investigating the solutions to some special matrix equations.. In continuation of Ran and Reurings, Nieto and Rodríguez-López6 who proved two very useful results and used them to solve some differential equations.
In 2012, Wardowski7 initiated the study of fixed points of a new type of contractive mappings in complete metric spaces. In 2014, Wardowski and Dung,8 proved fixed points of F -weak contractions on complete metric spaces. Acar et al.9 and Altun et al.10 gave Generalized multivalued F -contractions on complete metric spaces. In 2014, Minak et al.,11 proved Ćirić type generalized F -contractions on complete metric spaces and fixed point results. Paesano and Vetro12 gave the proof on Multi-valued F− contractions in 0-complete partial metric spaces with application to Volterra type integral equation. Piri and Kumam13 proved some fixed point theorems concerning F− contraction in complete metric spaces. Sawangsup and Sintunavarat14 proved the fixed point theorems for Fℜ contractions with applications to the solution of non-linear matrix equations, Tomar and Sharma15 proved some coincidence and common fixed point theorems concerning F− contraction and applications and Bashir16 proved the fixed point results of a generalized reversed F− contraction mapping and its application.
On the other hand, Alam and Imdad17 gave a generalizatiFor more res on of the Banach contraction principle in a complete metric space equipped with binary relation. Their results show that the contraction condition holds only for those elements linked with the binary relation, not for every pair of elements. Recently, Kannan’s and Reich’s18 fixed point theorems have been studied and extended in several directions, Karapinar19 modified the classical Kannan contraction phenomena to an interpolative Kannan contraction one to maximize the rate of convergence of an operator to a unique fixed point. However, by giving a counter-example, Karapinar and Agarwal20 pointed out a gap in the paper about the assumption of the fixed point being unique and came up with a corrected version. They provided a counter-example to verify that the fixed point need not be unique and invalidate the assumption of a unique fixed point. Since then, several results for variants of interpolative mapping proved for single and multivalued in various abstract spaces.
Further, Karapinar and Agarwal21 proved interpolative Rus-Reich-Ćirić type contractions via simulation functions. Errai et al.22 gave some new results of interpolative Hardy-Rogers and Reich-Rus-Ćirić type contraction in-metric spaces to prove the existence of the coincidence point. Mishra et al.23 proved the common fixed point theorems for interpolative Hardy-Rogers and Reich-Rus-Ćirić type contraction on quasi partial -metric space. Aydi et al.24 proved -interpolative Reich-Rus-Ćirić type contractions on metric spaces. Aydi et al.25 proved an interpolative Ćirić-Reich-Rus type contractions via the Branciari distance. Gautam et al.26 proved the fixed point of interpolative Rus-Reich-Ćirić contraction mapping on rectangular quasi-partial -metric space.
This manuscripts prove a fixed points theorem for relation-theoretic -contraction mappings via an arbitrary binary relation concept in Branciari distance metric space. In particular, we improve and extend the works due to Alam and Imdad,27 Ahmadullah et al.,28 Ahmadullah et al.,29 Eke et al.,30 Sawangsup and Sintunavarat,14 Aydi et al.31 and Karapinar et al.20 In doing so, we will generalize several other works in the literature having the same setting.
This section introduces some definitions, theorems and preliminary results, which will help develop the main result.
The concept of a Branciari distance space has been introduced by Brianciari where the triangular inequality is replaced by a quadrilateral one, which states as follows:
Definition 12. Let ℳ be a non-empty set. Suppose that the mapping d:ℳ×ℳ→[0,∞) be a function for all and all distinct points z,w∈ℳ , each distinct from μ and ν .
(i) d(μ,ν)≥0 and d(μ,ν)=0 if and only if μ=ν ;
(ii) d(μ,ν)=d(μ,ν) ;
(iii) d(μ,ν)≤d(μ,w)+d(w,z)+d(z,ν).
Then is called a Branciari distance and the pair (ℳ,d) is called a Branciari distance space.
Definition 22. Let (ℳ,d) be a metric space. A mapping Γ:ℳ→ℳ is said to be sequentially convergent if we have, for every sequence {νn} , if {Γνn} is convergence then {νn} also is convergence.ℳ is said to be subsequentially convergent if we have, for every sequence {νn} , if {Γνn} is convergence then {νn} has a convergent subsequence.
Definition 336. Let ( ℳ,d) be a Branciari distance space and {μn} be a sequence in ℳ .
(i)A sequence {μn} is converges to point μ⋆∈ℳ if limn→∞d(μn,μ⋆)=0
(ii) A sequence {μn} is said to be Cauchy if for every ϵ>0 , there exists a positive integer ℕ=ℕ(ϵ) such that d(μn,μm)<ϵ, for all n,m>ℕ .
(iii) We say that (ℳ,d) is complete if each Cauchy sequence in ℳ is convergent.
Lemma 136. Let (ℳ,d) be a Branciari distance space. A mapping Γ:ℳ→ℳ is continuous at μ⋆∈ℳ , if we have Γμn→Γμ⋆ or limn→∞d(Γμn,Γμ⋆)=0 , for any sequence {μn} in ℳ converges to μ⋆∈ℳ , that is μn→μ⋆ .
Proposition 137. Suppose μn is a Cauchy sequence in a Branciari distance space such that limn→∞d(μn,μ⋆)=d(μn,w⋆)=0, where μ⋆,w⋆∈ℳ . Then μ⋆=w⋆ .
Another noted attempt to extend the Banach contraction principle is essentially due to Wardowski.
The following explanations for developing the -contraction definition was obtained from Wardowski,7 Wardowski and Van Dung,8 and Cosentino et al.32
Let F:ℝ+→ℝ be a mapping satisfying:
(F1) F is strictly increasing, i.e. for all J,K∈ℝ+, J<K implies F(J)<F(K) ;
(F2) For each sequence {Jn}n∈ℕ of positive numbers,limn→∞Jn=0 if and only if limn→∞F(Jn)=−∞ ;
(F3) There exists z∈(0,1) satisfying limJn→0+JznF(Jn)=0
We denote the family of all functions F satisfying conditions (F1−F3) by ℱ . Some examples of functions F∈ℱ are:
F1(c)=lnc⇒d(Γμ,Γν)d(μ,ν)≤e−η;
F2(c)=c+lnc⇒d(Γμ,Γν)d(μ,ν)≤e−η+d(μ,ν)−d(Γμ,Γν) ;
F3(c)=−1√c⇒d(Γμ,Γν)d(μ,ν)≤1(1+η√d(μ,ν))2 ;
F4(c)=ln(c2+c)⇒d(Γμ,Γν)(1+d(Γμ,Γν))d(μ,ν)(1+d(μ,ν))≤e−η .
Wardowski introduced a generalization of the Banach contraction principle in metric spaces as follows:
Definition 47. Let (ℳ,d) be a metric space. A self-mapping Γ on ℳ is called an F -contraction mapping if there exists F∈ℱ and η∈ℝ+ such that for all μ,ν∈ℳ ,d(Γμ,Γν) >0⇒η+F(d(Γμ,Γν))≤F(d(μ,ν)).
Wardowski7 proved the following fixed point theorem:
Theorem 17. Let (ℳ,d) be a complete metric space and Γ:ℳ→ℳ be a F -contraction mapping. If there exist η>0 such that for all μ,ν∈ℳ, d(Γμ,Γν) >0 , implies η+F(d(Γμ,Γν))≤F(d(μ,ν)), then Γ has a unique fixed point.
Kannan1 proved the following theorem:
Theorem 21. Let (ℳ,d) be a complete metric space and a self-mapping Γ:ℳ→ℳ be a mapping such that d(Γμ,Γν)≤η{d(μ,Γμ)+d(ν,Γν)}, for all μ,ν∈ℳ and 0≤η≤12. The Γ has a unique fixed point δ∈ℳ and for any μ∈ℳ the sequence of iterate {Γnμ} converges to δ .
The following results for interpolative Kannan contraction have been proved in as follows:
Definition 519. Let (ℳ,d) be a metric space, the mapping Γ:ℳ→ℳ is said to be interpolative Kannan contraction mappings if d(Γμ,Γν)≤η[d(μ,Γμ)]δ.[d(ν,Γν)]1−δ, for all μ,ν∈ℳ with μ≠Γμ , where η∈[0,1) and δ∈(0,1) .
Theorem 319. Let (ℳ,d) be a complete metric space and Γ be an interpolative Kannan type contraction. Then Γ has a unique fixed point in ℳ .
In 2018, Karapinar et al.21 proved an interpolative Reich-Rus-Ćirić type contractions fixed point result on partial metric space as follows.
Theorem 421. Let (ℳ,d) be a complete metric space Γ:ℳ→ℳ . be a mapping such that p(Γμ,Γν)≤η[p(μ,ν)]δ.[p(μ,Γμ)]α.[p(ν,Γν)]1−α−δ, for all μ,ν∈ℳ∖Fix(Γ) where Fix(Γ)={μ∈ℳ,Γμ=μ} . Then Γ has a fixed point in ℳ .
Binary relation-theoretic in metric spaces
In this part, we will recall some definitions of relation theoretic notion related to binary relation with relevant relation-theoretical variants of some metrical concepts such as completeness and continuity which will be useful in developing our main results.
In the following discussion ℛ stands for a nonempty binary relation while ℕ0 denotes the set of whole numbers, i.e., ℕ∪{0} .
Definition 638. A binary relation on a non-empty set ℳ is defined as a subset of ℳ×ℳ , which will be denoted by ℛ . We say that μ relates to ν under ℛ iff (μ,ν)∈ℛ .
Definition 717. Let ℛ be a binary relation defined on a non-empty set ℳ and μ,ν∈ℳ . We say that μ and ν are ℛ -comparative if either (μ,ν)∈ℛ or (ν,μ)∈ℛ . We denote it by [μ,ν]∈ℛ .
Definition 838. Let ℛ be a binary relation defined on a non-empty set ℳ . Then the symmetric closure of ℛ is defined as the smallest symmetric relation containing ℛ(i.e.ℛs:=ℛ∪ℛ−1) , where ℛ−1={(μ,ν)∈ℳ2:(ν,μ)∈ℛ} .
Proposition 217. If ℛ is a binary relation defined on a non-empty set ℳ , then (μ,ν)∈ℛs⇔[μ,ν]∈ℛ.
Definition 917. Let ℛ be a binary relation defined on a non-empty set ℳ . Then a sequence {μn}⊂ℳ is called ℛ -preserving if (μn,μn+1)∈ℛ, ∀ n∈ℕ0 .
Definition 1017. Let ℳ be a non-empty set and Γ a self-mapping onℳ . A binary relation ℛ on X is called Γ -closed if for any μ,ν∈ℳ , (μ,ν)∈ℛ⇒(Γμ,Γν)∈ℛ.
Definition 1127. Let (ℳ,d) be a metric space and ℛ a binary relation on ℳ . We say that (ℳ,d) is ℛ -complete if every ℛ -preserving Cauchy sequence inℳ converges.
Definition 1217. Let (ℳ,d) be a metric space. A binary relation ℛ defined onℳ is called d-self closed if whenever {μn} is an ℛ -preserving sequence and μnd→μ , then there is a sub sequence {μnk} of {μn} with [μnk,μ]∈ℛ for all k∈ℕ0 .
Definition 1317. Letℳ be a non-empty set and ℛ a binary relation onℳ . A subset D ofℳ is called ℛ -directed if for each μ,ν∈D , there exists z inℳ such that (μ,z)∈ℛ and (ν,z)∈ℛ .
Definition 1439. Letℳ be a non-empty set and ℛ be a binary relation defined on a non-empty setℳ . Let k be a natural number, a path ℛ from μ to is a finite sequence {z0,z1,z2,...,zk}∈ℳ which satisfies the following conditions:
(vi) Let us denote γ(μ,ν,ℛ) : the collection of all paths {z0,z1,z2,...,zk} joining μ to ν in ℛ such that [zi,Γzi]∈ℛ for each i∈{1,2,3,...k−1} .
Further, we state some preliminary results which will be helpful to develop our main results.
Ahmadullah et al.33 proved the results in metric-like spaces as well as partial metric spaces equipped with a binary relation. Sawangsup and Sintunavarat14 by combining the concepts of Wardowski and proved the fixed point theorems for Fℜ -contractions in metric space with applications to the solution of non-linear matrix equations with binary relation as follows:
Theorem 534. Let (ℳ,d) be a complete metric space, ℛ a binary relation on ℳ and let Γ be a self-mapping on ℳ . Suppose that the following conditions hold:
(i) ℳ(Γ,ℛ) is non-empty,
(ii)ℛ is Γ -closed,
(iii) either Γ is continuous or ℛ is G -self-closed,
(iv)there exists F∈ℱ and η∈ℝ+ such that for all μ,ν∈ℳ with (μ,ν∈ℛ) ,d(Γμ,Γν)>0⇒η+F(d(Γμ,Γν))≤F(d(μ,ν)).
Then Γ has a fixed point. Moreover, for each x0∈ℳ(Γ,ℛ) the Picard sequence {Γnx0} is convergent to the fixed point Γ .
Now, we prove the main results using interpolative Reich-Rus-Ćirić- -contraction mapping concepts via binary relation in generalized metric spaces.
Theorem 6. Let (ℳ,d) be a complete metric space, ℛ a binary relation on ℳ and let Γ be an interpolative Reich-Rus-Ćirić type contractions mapping on ℳ . Suppose that the following conditions hold:
(i)(ℳ,d) is Γ complete,
(ii)ℳ(Γ,ℛ) is non-empty,
(iii)ℛ is Γ -closed,
(iv) the sequence {μn} is ℛ -preserving,
(v) either Γ is continuous or ℛ is d-self closed,
(vi) there exists a constant η>0 such that ∀μ,ν∈ℳ with (μ,ν∈ℛ) η+F(d(Γμ,Γν))≤F(ℳℛ(μ,ν)), μ,ν∈ℳ∖Fix(Γ) where for all where Fix(Γ)={μ∈ℳ,Γμ=μ} . Then Γ has a fixed point. Also, if Γ is subsequentially convergent then for every μn−1∈ℳ the sequence of iterate {Γnμn−1} converges to this fixed point. Moreover, if
(vii)γ(μ,ν,ℛs) is non-empty, for each μ,ν∈ℳ . Then Γ has a unique fixed point.
Proof. Assume x0 be an arbitrary point in ℳ(Γ,ℛ) . We construct a sequence {μn} of Picard iterates such that μn=Γnμ0=Γμn−1 for all n∈ℕ . By condition (iii) of Theorem 6, we have (μ0,Γμ0)∈ℛ and ℛ is Γ -closed, therefore
(Γμn−1,Γn+1μn−1),(Γn+1μn−1,Γn+2μn−1),…,(Γnμn−1,Γn+2μn−1) .
Using $(???)$, we note that
(Γnμn−1,Γn+1μn−1)∈ℛ,
∀n∈ℕ0. Therefore the sequence {μn} is ℛ -preserving.
If there exists n such that μn=μn+1 , then μn is a fixed pint of Γ . The proof is completed. For that case, we assume that μn≠μn+1 for each n≥0 . Therefore
limn→∞d(μn,μn+1)=0.
To show this, let μ=μn−1 and ν=μn , using $(???)$ for all n∈ℕ0 , we deduce that
η+F(d(Γμn−1,Γμn))≤F(ℳℛ(μn−1,μn)),
where
ℳℛ(μn−1,μn)=[d(μn−1,μn)]δ.[d(μn−1,Γμn−1)]α.[d(μn,Γμn)]1−α−δ,≤[d(μn−1,μn)]δ.[d(μn−1,μn)]α.[d(μn,μn+1)]1−α−δ,=[d(μn−1,μn)]α+δ.[d(μn,μn+1)]1−α−δ.
Taking $(???)$ into $(???)$, we obtain
η+F(d(Γμn−1,Γμn))≤F([d(μn−1,μn)]α+δ.[d(μn,μn+1)]1−α−δ).
By the continuity property of , and $(???)$, we get
d(Γμn−1,Γμn)≤[d(μn−1,μn)]α+δ.[d(μn,μn+1)]1−α−δ,(d(μn,μn+1))1−(1−α−δ)≤[d(μn−1,μn)]α+δ,(d(μn,μn+1))α+δ≤[d(μn−1,μn)]α+δ.
So, we conclude that
d(μn,μn+1)≤d(μn−1,μn),
for all n≥1 .
Consequently, we have
η+F(d(μn,μn+1))≤F(d(μn−1,μn)).
Equivalent to
F(d(μn,μn+1))≤F(d(μn−1,μn))−η.
Similar, let μ=μn , ν=μn+1 , using $(???)$ and $(???)$ for all n∈ℕ0 , we get
F(d(μn+1,μn+2))≤F(d(μn−1,μn))−2η.
Proceeding this way, by induction we deduce
F(d(μn,μn+1))≤F(d(μn−1,μn))−nη,∀n≥1.
That is d(μn−1,μn) is non-increasing sequence with non-negative terms. We denote Jn=d(μn,μn+1) , for all n∈ℕ0 . Since Γ is an F - ℛ -interpolative contraction mapping.
From $(???)$, we obtain
F(Jn)≤F(Jn−1)−η≤F(Jn−2)−2η≤…≤F(J0)−nη,
for all n∈ℕ0 .
By (F2) , we have
limn→∞Jn=0.
If and only if
limn→∞F(Jn)=−∞.
From (F3) and $(???)$, there exists z∈(0,1) such that
JznF(Jn)≤Jzn(F(Jn−1)−η)≤…≤Jzn(F(J0)−nη),JznF(Jn)≤Jzn(F(J0)−nη)≤0,JznF(Jn)≤JznF(J0)−Jznnη)≤0,JznF(Jn)−JznF(J0)≤−Jznnη≤0,Jzn(F(Jn)−F(J0))≤−Jznnη≤0.
Letting n→∞ in $(???)$, we obtain that
limn→∞Jznn=0.
Now, from $(???)$ there exist n1∈ℕ0 such that Jznn≤1 for all n≥n1 .
Consequently, we have that
Jznn≤1,Jzn≤1n,Jn≤1n1z,Jn≤n−1z.
Therefore, ∑∞n=0d(μn,μn+1)=0 converges.
Next, we claim that {μn} is Cauchy sequence, that is, limn→∞d(μn,μm)=0 ∀n,m∈ℕ0 such that m≥n , by using the rectangular property we have
d(μn,μm)≤d(μn,μn+1)+d(μn+1,μn+2)+…+d(μm−1,μm),≤Jn+Jn+1+Jn+2+…+Jm−1,=∑m−1i=nJi,≤∑m−1i=nn−1z.
Since ∑m−1i=nn−1z<∞ , we get that {μn} is a Cauchy sequence in ℳ . Since (ℳ,d) is complete, there exists μ⋆∈ℳ such that
d(μn,μ⋆)=limn→∞d(μn,μ⋆)=0.
Now, by the continuity of Γ , we get Γμ⋆=μ⋆ . We show that μ⋆ is a fixed point of Γ . Assume that Γμ⋆≠μ⋆ such that Γμn≠μn ∀n≥ℕ0 . By letting μ=μn and ν=μ⋆ in $(???)$, we obtain
η+F(d(Γμn,Γμ⋆))≤F(ℳℛ(μn,μ⋆)),
where
ℳℛ(μn,μ⋆)=[d(μn,μ⋆)]δ.[d(μn,Γμn)]α.[d(μ⋆,Γμ⋆)]1−α−δ,≤[d(μ⋆,μ⋆)]δ.[d(μ⋆,μ⋆)]α.[d(μ⋆,μ⋆)]1−α−δ,=[d(μ⋆,μ⋆)](α+δ)+(1−α−δ),=d(μ⋆,μ⋆).
Taking $(???)$ into $(???)$, we get
η+F(d(Γμ⋆,Γμ⋆))≤F(d(μ⋆,μ⋆)),η+F(0)≤F(0),η≤0,
which is a contradiction. Hence, d(μ⋆,Γμ⋆)=0 therefore μ⋆=Γμ⋆ , which shows that μ⋆ is a fixed point of Γ . Also Γ is subsequentially convergent on ℳ . To observe this, let μ=μnk−1 and ν=μnk , using $(???)$ we obtain
η+F(d(Γμnk−1,Γμnk))≤F(ℳℛ(μnk−1,μnk)),
where
ℳℛ(μnk−1,μnk)=[d(μnk−1,μnk)]δ.[d(μnk−1,Γμnk−1)]α.[d(μnk,Γμnk)]1−α−δ,
≤[d(μnk−1,μnk)]δ.[d(μnk−1,μnk)]α.[d(μnk,μnk+1)]1−α−δ,
= [d(μnk−1,μnk)]α+δ.[d(μnk,μnk+1)]1−α−δ.
Using $(???)$ in $(???)$, we get
η+F(d(Γμnk−1,Γμnk))≤F([d(μnk−1,μnk)]α+δ.[d(μnk,μnk+1)]1−α−δ).
By the property of F and F1 with $(???)$, we get
d(Γμnk−1,Γμnk)≤[d(μnk−1,μnk)]α+δ.[d(μnk,μnk+1)]1−α−δ,(d(μnk,μnk+1))1−(1−α−δ)≤[d(μnk−1,μnk)]α+δ,(d(μnk,μnk+1))α+δ≤[d(μnk−1,μnk)]α+δ,
which is equivalent to
F(d(μnk,μnk+1))≤F(d(μnk−1,μnk))−η.
Due to continuity of Γ , it implies that
limn→∞Γμnk=Γμ⋆=μ⋆.
This shows that Γ is subsequentially convergent.
Consider the hypothesis in Theorem 6, we prove assertion (vii) as follows: we observe that ℳ(Γ,ℛ) is non-empty, so let us take a pair of elements say (μ⋆,w⋆) in ℳ(Γ,ℛ) such that
Γμ=μ⋆,Γν=w⋆.
Next, we claim that μ⋆≠w⋆ . By the above equalities, there exists a S-path (say,z0,z1,z2,...,zl) of length in ℛs from Γμ to Γν , with
Γz0=Γμ,Γzl=Γν,
such that
[Γzi,Γzi+1]∈ℛs⊆ℛ,
for all i∈0,1,2,3,...l−1 .
Define two constant sequences such that
z0n=μ and zln=ν
By using ([equation 4.20]), for all n∈ℕ , we have
Γz0n=Γμ=μ⋆,Γzln=Γν=w⋆.
By usual substitution for zi0=zi for each i∈0,1,2,...l , that is
z10=z1,z20=z2,z30=z3,z40=z4,zl−10=zl−1.
Thus we construct a sequence
{z1n},{z2n},{z3n},…,{zin}∈ℳ.
Corresponding to each zi , we have [Γzi0,Γzi1]∈ℛ from ([equation 4.20]), ([equation 4.21]) and ℛ is Γ -closed, we get
limn→∞d(Γzin,Γzi+1n)=0,
for each i∈1,2,3,...l−1 and for all n∈ℕ .
Define din=d(Γzin,Γzi+1n) for each i∈0,1,2,3,...l−1 and for all n∈ℕ . We assert that, limn→∞din>0 .
Since [Γzin,Γzi+1n]∈ℛ , either [Γzin,Γzi+1n]∈ℛ or [Γzi+1n,Γzin]∈ℛ .
If [Γzin,Γzi+1n]∈ℛ , for μ=zin and ν=zi+1n . Then applying the condition $(???)$, we have
η+F(d(Γzin,Γzi+1n))≤F(ℳℛ(zin,zi+1n)),
where
ℳℛ(zin,zi+1n) = [d(zin,zi+1n)]δ.[d(zin,Γzin)]α.[d(zi+1n,Γzi+1n)]1−α−δ,
≤ [d(zin,zi+1n)]δ.[d(zin,zi+1n)]α.[d(zi+1n,zi+2n)]1−α−δ,
= [d(zin,zi+1n)]α+δ.[d(zi+1n,zi+2n)]1−α−δ.
Substituting $(???)$ in $(???)$, we get
η+F(d(zi+1n,zi+2n))≤F([d(zin,zi+1n)]α+δ.[d(zi+1n,zi+2n)]1−α−δ).
By the property of F , we have
d(zi+1n,zi+2n)≤[d(zin,zi+1n)]α+δ.[d(zi+1n,zi+2n)]1−α−δ,d(zi+1n,zi+2n)1−(1−α−δ)≤[d(zin,zi+1n)]α+δ,d(zi+1n,zi+2n)α+δ≤[d(zin,zi+1n)]α+δ,d(zi+1n,zi+2n)≤d(zin,zi+1n).
Which is equivalent to
η+F(d(zi+1n,zi+2n))≤F(d(zin,zi+1n)),F(d(zi+1n,zi+2n))≤F(d(zin,zi+1n))−η.
Taking lim as i→∞ and using limi→∞din=d , we get
d(Γzin,Γzi+1n)≤0.
Implies that
η≤0,
which is a contradiction and hence
limi→∞din=d=0
The same for rectangular property (iii), if (Γzin,Γzi+1n)∈ℛ , we have
limi→∞din=limi→∞d(Γzin,Γzi+1n)=0,
for i∈0,1,2,...l−1 .
Using ([equation 4.21]),limi→∞din=0 and (iii ), we have
d(μ⋆,w⋆)=d(zin,zjn)≤∑l−1i=0d(zin,zi+1n),≤∑l−1i=0din,→0 as n→∞.
So that
d(μ⋆,w⋆)=0⇒μ⋆=w⋆.
Therefore
Γμ⋆=Γw⋆,
which is a contradiction. Thus μ⋆ is a unique fixed point of Γ . Thus the proof is completed.
Due to the generalization of Theorem 6, we can deduce the corollary as follows:
Corollary 1. Let (ℳ,d) be a complete metric space and let Γ:ℳ→ℳ be F -interpolative type mapping such that the following hypothesis hold:
(i)
ℳ is Γ is closed in (ℳ,d) ,
(ii)
there exists a constant η∈[0,1) such that
η+F(d(Γμ,Γν))≤F([d(μ,Γμ)]δ.[d(ν,Γν)]1−δ),
for all μ,ν∈ℳ with μ≠Γμ , where η∈[0,1) and δ∈(0,1) .
Proof. The proof of the above corollary follows similar steps of Theorem 6. Therefore, the proof is completed.
Next, we give the following similar example from Moradi and Alimohammadi4 for illustration of the hypothesis of Theorem 6.
Example 1. Consider ℳ={0}∪{0,1,12,13} and d be a Euclidean metric on ℳ . Then (ℳ,d) is a complete metric space. The mapping Γ:ℳ→ℳ be determined as Γ(0)=0 ∀n=0,Γ(μ)=1μn+1,∀n≥1. Define a binary relation ℛ={(μ,ν)∈ ℝ2},ℛ∈ℝ2 and ℛ={(0,1),(0,12),(0,13),(1,13),(1,12),(12,13)} on ℳ. Then ℳ is ℛ -complete.
We claim that ℳ is not either ℛ -complete or Γ -closed. To verify this, we show that F3(c)=−1√c⇒d(Γμ,Γν)ℳℛ(μ,ν)≤1(1+η√ℳℛ(μ,ν))2 satisfy all the hypothesis of Theorem 6.
We complete the following metrics Using all of the above equalities, we obtain ℳℛ(μ,ν)=[|μ−ν|]δ.[|μn+1+μ−1μn+1|]α.[|νn+1+ν−1νn+1|]1−α−δ, By substituting $(???)$ in $(???)$, we obtain |μn−νn(μn+1)(νn+1)|[|μ−ν|]δ.[|μn+1+μ−1μn+1|]α.[|νn+1+ν−1νn+1|]1−α−δ≤1(1+η√ℳℛ(μ,ν))2. If we take δ=0.2,α=0.5,η=13 and n=1 in the above inequality, for all (μ,ν)∈ℛ , such that δ+α≤1 . We conclude that ℳ is either ℛ -complete or Γ -closed. Which is a contradiction to our claim. Hence, all the hypotheses of Theorem 6 are satisfied.
An application to non linear matrix equations
In this section, we prove the existence of the solution for the nonlinear matrix equation. We use one application to utilize the results obtained in Theorem 6, where a fixed point solution is applied to complete the Branciari distance. We refer to the study of the nonlinear matrix equation from Ran and Reurings5 who proved a fixed point theorem in partially ordered sets and some applications to matrix equations. The Hermitian solution of the equation X=Q+NX−1N* is the matrix equation arising from the Gaussian process. The equation admits both definite positive solution and definite negative solution if and only if N is non-singular. If N is singular, no definite negative solution exists. Nonlinear matrix equations play an important role in several problems that arise in the analysis of control theory and system theory.
The main concern of this section is to apply Theorem 6 to study the following nonlinear matrix equations, which are motivated by Jain et al.,34 Lim et al.,35 Sawangsup and Sintunavara,14 Ran and Reurings5 and several others.
μ=Q+∑ni=1N*iΓ(μ)Ni,Q=μ−N*1Γ(μ)N1−…−N*nΓ(μ)Nn,
where ℋ(n) is a set of n×n Hermitian matrices, p(n) is a set of n×n positive definite matrices and p(n)⊆ℋ(n) ,Q ∈p(n) is a Hermitian positive definite matrix,Ni is n×n matrices and Γ;p(n)→p(n) is a continuous order-preserving map such that Γ(0)=0 .
The set ℋ(n) equipped with the trace norm is a complete metric space and partially ordered with partial ordering , where equivalently .
We use the following lemmas from Ran and Reurings5 that will be useful for developing our results.
Lemma 25. If are matrices, then
Lemma 35. If , then .
Now, we prove a fixed point for self-mappings for the following nonlinear matrix equation in Branciari distance.
where Q , is matrices, stands for conjugate transpose of and is a continuous order-preserving map such that .
Theorem 7. Consider the class of nonlinear matrix Equation [equation 3.5] and suppose the following condition holds.
(i)
there exists with
(ii)
for all ,
(iii)
There exist for which and such that for all we have and
(iv)
there exist and such that where Then, the non linear matrix equation $(\ref{equation 3.5})$ has a solution in
Proof. Define by
for all . Then the fixed point of the mapping is a solution of the matrix equation $(\ref{equation 3.5})$.
The Branciari metric is defined by
Let be well defined on and -closed. For with , then . We claim that is not an - -contraction mapping with respect to and and by using we get
which is a contradiction. Hence is a contraction. Therefore, from , we have . Thus, by using Theorem 6 we conclude that has a unique fixed point in and .
The new concept of relation-theoretic -interpolative mapping endowed with binary relation in Branciari Distance in metric spaces has been introduced. In particular, we improved and extended the works due to Alam and Imdad17, Ahmadullah et al.28 , Ahmadullah et al29. , Eke et al.30 , Sawangsup and Sintunavarat14 ,Aydi et al.24 and Karapinar et al20. . In doing so, we generalized several other works in the literature having the same setting. Henceforth, the results obtained will be verified with the help of illustrative examples. Also, we demonstrate the results with an application in matrix equations
Conflict of interest: Authors declare that they have no conflicts of interest.
Research involving human participants and/or animals: The author declares that there are no human participants and/or animals involved in this research.
Authors declare that there is no funding available for this research.
ADD THE FOLLOWING REFERENCES
©2023 Wangwe. This is an open access article distributed under the terms of the, which permits unrestricted use, distribution, and build upon your work non-commercially.