Skip to main content

Erratum to: Nonlinear quasi-contractions in non-normal cone metric spaces

The Original Article was published on 28 July 2014

Abstract

In the note we correct some errors that appeared in the article (Jiang and Li in Fixed Point Theory Appl. 2014:165, 2014).

MSC:06A07, 47H10.

Correction

Upon critical examination of the main results and their proofs in [1], we note some critical errors under the conditions of the main theorem and its proof in our article [1].

In this note, we would like to supplement some essential conditions, which will ensure that the mapping B is well defined, to achieve our claim.

The following theorem is a slight modification of [[1], Theorem 1].

Theorem 1 Let (X,d) be a complete cone metric space over a solid cone P of a Banach space (E,) and T:XX a quasi-contraction (i.e., there exists a mapping A:PP such that

d(Tx,Ty)Au,x,yX,
(1)

where u{d(x,y),d(x,Tx),d(y,Ty),d(x,Ty),d(y,Tx)}). Assume that A:PP is a nondecreasing, continuous and subadditive (i.e., A(u+v)Au+Av for each u,vP) mapping with Aθ=θ such that

i = 0 A i u <,uP.
(2)

If B is continuous at θ, where Bu= i = 0 A i u for each uP, then T has a unique fixed point x X, and for each x 0 X, the Picard iterative sequence { x n } converges to x , where x n = T n x 0 for each n.

Remark 1 In the case that the normed vector space (E,) is complete, if (2) holds then the mapping B is well defined. In fact, fix uP, let s n = i = 0 n A i u and S n = i = 0 n A i u. By (2), we get lim n S n = i = 0 A i u and hence { S n } is a Cauchy sequence of reals. Note that s m s n = i = n + 1 m A i u i = n + 1 m A i u= S m S n for each m>n, then { s n } is a Cauchy sequence in E. Moreover, by the completeness of E, { s n } is convergent. This implies that Bu= lim n i = 0 n A i u for each u, i.e., B is well defined. However, in [[1], Theorem 1] the normed vector space E is not assumed to be complete, then { s n } may not be convergent, and consequently, B may be not meaningful.

Remark 2 (i) In [1] the authors claim that (see (4) in [1])

BA=AB,
(3)

which plays an important role in the proof of [[1], Theorem 1]. However, if A is a nonlinear mapping, the above claim may not hold. For example, let E=P=[0,a] and A(t)= t 2 for each tP, where 0<a<1. It is clear that A:PP is nonlinear. Note that A 0 (t)=t and A i (t)= t 2 i (i=1,2,3,), then i = 0 A i (t)=t+ i = 1 t 2 i t+ i = 1 t 2 i =t+ t 2 1 t 2 for each t[0,a], and hence i = 0 A i (t) is convergent for each t[0,a]. This implies that the function B(t) is well defined, where B(t)= i = 0 A i (t) for each t[0,a]. For each t[0,a], we have AB(t)= t + i = 1 t 2 i and BA(t)= i = 1 t 2 i . Suppose that there exists t 0 (0, 2 2 ] such that AB( t 0 )=BA( t 0 ), and set b=BA( t 0 ). Then we have 0<b t 0 2 1 t 0 2 and t 0 +b= b 2 . Solve the equation t 0 +b= b 2 , then b= 1 + 1 + 4 t 0 2 by b>0. Thus we get 1< 1 + 1 + 4 t 0 2 =b t 0 2 1 t 0 2 1, a contradiction. Hence BA(t)AB(t) for each t(0, 2 2 ]. This shows that (3) does not hold.

(ii) Note that A is not confined to a linear mapping in [[1], Theorem 1], then from (i) we know that (3) may not hold, and consequently, the proof of [[1], Theorem 1] is not finished yet. In order to complete its proof, we add the continuity of A to Theorem 1.

(iii) Suppose that E is a Banach space and A is a continuous and subadditive mapping such that (2) is satisfied, then by Remark 1 we get

A B u = A ( lim n i = 0 n A i u ) = lim n A ( i = 0 n A i u ) lim n ( i = 0 n A i + 1 u ) = i = 1 A i u = B A u , u P .
(4)

In what follows, we shall complete the proof of Theorem 1 by using (4) instead of (3). Since there are too many changes required for the proof of [[1], Theorem 1], we present the full proof of Theorem 1 as follows.

Proof of Theorem 1 It follows from (2) and Remark 1 that the mapping B is well defined. Clearly, B is a nondecreasing and subadditive mappings with B(P)P and Bθ=θ since A is nondecreasing and subadditive, A(P)(P) and Aθ=θ. We claim that for each n1,

d( x i , x j )BAd( x 0 , x 1 ),1i,jn.
(5)

In the following we shall show this claim by induction.

If n=1, then i=j=1, and so the claim is trivial.

Assume that (5) holds for n. To prove (5) holds for n+1, it suffices to show

d( x i 0 , x n + 1 )BAd( x 0 , x 1 ),1 i 0 n.
(6)

By (1),

d( x i 0 , x n + 1 )Au,
(7)

where

u { d ( x i 0 1 , x n ) , d ( x i 0 1 , x i 0 ) , d ( x n , x n + 1 ) , d ( x i 0 1 , x n + 1 ) , d ( x n , x i 0 ) } .

Consider the case that i 0 =1.

If u=d( x 0 , x n ), then by the triangle inequality, the nondecreasing property, subadditivity of A, the definition of B, (4), (5), and (7)

d ( x i 0 , x n + 1 ) A d ( x 0 , x n ) A [ d ( x 0 , x 1 ) + d ( x 1 , x n ) ] A [ d ( x 0 , x 1 ) + B A d ( x 0 , x 1 ) ] A d ( x 0 , x 1 ) + A B A d ( x 0 , x 1 ) A d ( x 0 , x 1 ) + B A 2 d ( x 0 , x 1 ) = A d ( x 0 , x 1 ) + i = 2 A i d ( x 0 , x 1 ) = i = 1 A i d ( x 0 , x 1 ) = B A d ( x 0 , x 1 ) ,

i.e., (6) holds.

If u=d( x 0 , x 1 ), then by the definition of B and (7)

d( x i 0 , x n + 1 )Ad( x 0 , x 1 ) i = 1 A i d( x 0 , x 1 )=BAd( x 0 , x 1 ),

i.e., (6) holds.

If u=d( x 0 , x n + 1 ), then by the triangle inequality, the nondecreasing property and subadditivity of A, and (7)

d ( x i 0 , x n + 1 ) A d ( x 0 , x n + 1 ) A [ d ( x 0 , x 1 ) + d ( x i 0 , x n + 1 ) ] A d ( x 0 , x 1 ) + A d ( x i 0 , x n + 1 ) .

Acting on the above inequality with B, by the nondecreasing property and subadditivity of B

B d ( x i 0 , x n + 1 ) B [ A d ( x 0 , x 1 ) + A d ( x i 0 , x n + 1 ) ] B A d ( x 0 , x 1 ) + B A d ( x i 0 , x n + 1 ) ,

which together with the definition of B implies that

d( x i 0 , x n + 1 )=Bd( x i 0 , x n + 1 )BAd( x i 0 , x n + 1 )BAd( x i 0 , x n + 1 ),

i.e., (6) holds.

If u=d( x n , x i 0 ), then by the definition and the nondecreasing property of A, (4), (5), and (7)

d ( x i 0 , x n + 1 ) A d ( x i 0 , x n ) A B A d ( x 0 , x 1 ) B A 2 d ( x 0 , x 1 ) = i = 2 A i d ( x 0 , x 1 ) i = 1 A i d ( x 0 , x 1 ) = B A d ( x 0 , x 1 ) ,

i.e., (6) holds.

If u=d( x n , x n + 1 ), we set i 1 =n1, and then by (7)

d( x i 0 , x n + 1 )Ad( x i 1 , x n + 1 ).
(8)

Consider the case that 2 i 0 n.

If u=d( x i 0 1 , x n ), or u=d( x i 0 1 , x i 0 ), or d( x n , x i 0 ), then by the definition and the nondecreasing property of A, (4), (5), and (7)

d ( x i 0 , x n + 1 ) A u A B A d ( x 0 , x 1 ) B A 2 d ( x 0 , x 1 ) = i = 2 A i d ( x 0 , x 1 ) i = 1 A i d ( x 0 , x 1 ) = B A d ( x 0 , x 1 ) ,

i.e., (6) holds.

If u=d( x n , x n + 1 ), or u=d( x i 0 1 , x n + 1 ), we set i 1 =n, or i 1 = i 0 11, respectively, and then (8) follows.

From the above discussions of both cases, we find the result that either (6) holds, and so the proof of our claim is complete, or there exists i 1 {1,2,,n} such that (8) holds. For the latter situation, continuing in a similar way, it will be found as a result that either

d( x i 1 , x n + 1 )BAd( x 0 , x 1 ),

which together with the definition and the nondecreasing property of A, (4), and (8), forces that

d ( x i 0 , x n + 1 ) A B A d ( x 0 , x 1 ) B A 2 d ( x 0 , x 1 ) = i = 2 A i d ( x 0 , x 1 ) i = 1 A i d ( x 0 , x 1 ) = B A d ( x 0 , x 1 ) ,

i.e., (6) holds, and so the proof of our claim is complete; or there exists i 2 {1,2,,n} such that

d( x i 1 , x n + 1 )Ad( x i 2 , x n + 1 ).

If the above procedure ends by the k th step with kn1, that is, there exist k+1 integers i 0 , i 1 ,, i k {1,2,,n} such that

d ( x i 0 , x n + 1 ) A d ( x i 1 , x n + 1 ) , d ( x i 1 , x n + 1 ) A d ( x i 2 , x n + 1 ) , , d ( x i k 1 , x n + 1 ) A d ( x i k , x n + 1 ) , d ( x i k , x n + 1 ) B A d ( x 0 , x 1 ) ,

then by the nondecreasing property of A and (4)

d ( x i 0 , x n + 1 ) A k B A d ( x 0 , x 1 ) B A k + 1 d ( x 0 , x 1 ) = i = k + 1 A i d ( x 0 , x 1 ) i = 1 A i d ( x 0 , x 1 ) = B A d ( x 0 , x 1 ) ,

i.e. (6) holds, and so the proof of our claim is complete.

If the above procedure continues more than n steps, then there exist n+1 integers i 0 , i 1 , i n {1,2,,n} such that

d ( x i 0 , x n + 1 ) A d ( x i 1 , x n + 1 ) , d ( x i 1 , x n + 1 ) A d ( x i 2 , x n + 1 ) , , d ( x i n 1 , x n + 1 ) A d ( x i n , x n + 1 ) .
(9)

It is clear that i 0 , i 1 , i n {1,2,,n} implies there exist two integers k,l{0,1,2,,n} with k<l such that i k = i l , then by the nondecreasing property of A and (9)

d( x i k , x n + 1 ) A l k d( x i l , x n + 1 )= A l k d( x i k , x n + 1 ).
(10)

Acting on (10) with B, by the nondecreasing property of B we get

Bd( x i k , x n + 1 )B A l k d( x i k , x n + 1 ),

which together with the definition of B implies that

d( x i k , x n + 1 ) j = 0 l k 1 A j d( x i k , x n + 1 )=Bd( x i k , x n + 1 )B A l k d( x i k , x n + 1 )θBAd( x 0 , x 1 ),

i.e., (6) holds. The proof of our claim is complete.

Note that B and A are nondecreasing and continuous at θ, Bθ=θ and Aθ=θ, then it follows from Lemma 3 of [1] that for each { u n }P,

B u n w θ,A u n w θ,BA u n w θ,
(11)

provided that u n w θ. By (2), we get

lim n A n u =0,uP.
(12)

Then in analogy to the proof of [[1], Theorem 1], by (5), (11), (12) we can show that

d( x m , x n ) w θ(n>m),
(13)

and there exists some x X such that

d ( x n , x ) w θ(n).
(14)

By (1),

d ( T x , x ) d ( x n + 1 , T x ) +d ( x n + 1 , x ) Au+d ( x n + 1 , x ) ,n,
(15)

where u{d( x n , x ),d( x n , x n + 1 ),d( x ,T x ),d( x n ,T x ),d( x , x n + 1 )}.

If u=d( x n , x ), or u=d( x n , x n + 1 ), or u=d( x , x n + 1 ), then it follows from (11), (13), and (14) that Au w θ and hence d(T x , x )=θ by (15).

If u=d( x ,T x ), then by (15)

d ( x , T x ) Ad ( x , T x ) +d ( x n + 1 , x ) ,n.

Acting on the above inequality with B, by the nondecreasing and subadditivity of B we get

Bd ( x , T x ) B [ A d ( x , T x ) + d ( x n + 1 , x ) ] BAd ( x , T x ) +Bd ( x n + 1 , x ) ,n,

which together with the definition of B implies that

d ( x , T x ) =Bd ( x , T x ) BAd ( x , T x ) Bd ( x n + 1 , x ) ,n,

and hence d( x ,T x )=θ since Bd( x n + 1 , x ) w θ by (11) and (14).

If u=d( x n ,T x ), then, by the triangle inequality, the nondecreasing property, and subadditivity of A and (15), we get

d ( T x , x ) d ( x n + 1 , x ) + A d ( x n , T x ) d ( x n + 1 , x ) + A [ d ( x n , x ) + d ( x , T x ) ] d ( x n + 1 , x ) + A d ( x n , x ) + A d ( x , T x ) , n .

Acting on the above inequality with B, then by the nondecreasing property and subadditivity of B

B d ( x , T x ) B [ d ( x n + 1 , x ) + A d ( x n , x ) + A d ( x , T x ) ] B d ( x n + 1 , x ) + B A d ( x n , x ) + B A d ( x , T x ) , n ,

which together with the definition of B implies that

d ( x , T x ) =Bd ( x , T x ) BAd ( x , T x ) Bd ( x n + 1 , x ) +BAd ( x n , x ) ,n,

and hence d( x ,T x )=θ since Bd( x n + 1 , x ) w θ and BAd( x n , x ) w θ by (11) and (14). This shows that x is a fixed point of T.

If x is another fixed point of T, then by (1)

d ( x , x ) =d ( T x , T x ) Au,

where u{d(x, x ),d(x,Tx),d( x ,T x ),d(x,T x ),d( x ,Tx)}. If u=d(x,Tx), or u=d( x ,T x ), then u=θ, and hence d(x, x )=θ since Aθ=θ. If u=d(x, x ), or u=d(x,T x ) or u=d( x ,Tx), then we must have u=d(x, x ), and hence d(x, x )Ad(x, x ). Acting on it with B, by the nondecreasing property of B we get Bd(x, x )BAd(x, x ). Moreover, by the definition of B, we have d(x, x )=Bd(x, x )BAd(x, x )θ and hence d(x, x )=θ. This shows x is the unique fixed point of T. The proof is complete. □

References

  1. Jiang S, Li Z: Nonlinear quasi-contraction in non-normal cone metric spaces. Fixed Point Theory Appl. 2014., 2014: Article ID 165

    Google Scholar 

Download references

Acknowledgements

The work was supported by Natural Science Foundation of China (11161022, 11461029, 71462015), Natural Science Foundation of Jiangxi Province (20122BAB201015, 20142BCB23013, 20143ACB21012), Educational Department of Jiangxi Province (KJLD14034), and Program for Excellent Youth Talents of JXUFE (201201).

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Zhilong Li.

Additional information

The online version of the original article can be found at 10.1186/1687-1812-2014-165

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0), which permits use, duplication, adaptation, distribution, and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Li, Z., Jiang, S. Erratum to: Nonlinear quasi-contractions in non-normal cone metric spaces. Fixed Point Theory Appl 2014, 196 (2014). https://doi.org/10.1186/1687-1812-2014-196

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/1687-1812-2014-196

Keywords

  • nonlinear quasi-contraction
  • cone metric space