• No results found

On some new matrix transformations

N/A
N/A
Protected

Academic year: 2020

Share "On some new matrix transformations"

Copied!
7
0
0

Loading.... (view fulltext now)

Full text

(1)

R E S E A R C H

Open Access

On some new matrix transformations

Rahmet Sava¸s

1*

and Ekrem Sava¸s

2

*Correspondence:

[email protected] 1Department of Mathematics, Medeniyet University, Istanbul, Turkey

Full list of author information is available at the end of the article

Abstract

In this paper, we characterize some matrix classes (ω(p,s),Vσλ), (ωp(s),Vσλ) and (ωp(s),Vσλ)regunder appropriate conditions.

1 Introduction

Letwdenote the set of all real and complex sequencesx= (xk). Byl∞andc, we denote the Banach spaces of bounded and convergent sequencesx= (xk) normed byx=supk|xk|, respectively. A linear functionalLonl∞is said to be a Banach limit [] if it has the following properties:

() L(x)≥ifn≥(i.e.,xn≥for alln),

() L(e) = , wheree= (, , . . .),

() L(Dx) =L(x), where the shift operatorDis defined byD(xn) ={xn+}.

LetBbe the set of all Banach limits onl∞. A sequencex∞is said to be almost con-vergent if all Banach limits ofxcoincide. Letcˆdenote the space of the almost convergent sequences. Lorentz [] has shown that

ˆ

c=

xl∞:lim

m dm,n(x) exists, uniformly inn

,

where

dm,n(x) =

xn+xn++xn++· · ·+xn+m

m+  .

The study of regular, conservative, coercive and multiplicative matrices is important in the theory of summability. In [], King used the concept of the almost convergence of a sequence introduced by Lorentz to define more general classes of matrices than those of regular and conservative ones.

In [], Schaefer defined the concepts ofσ-conservative,σ-regular andσ-coercive ma-trices and characterized the matrix classes (c,), (c,)reg and (l∞,), where

de-notes the set of all bound sequences, all of whose invariant means (orσ-means) are equal. In [], Mursaleen characterized the classes (c(p),), (c(p),)reg and (l∞(p),) of

ma-trices, which generalized the results due to Schaefer []. In [], Mohiuddine and Aiyup defined the spaceω(p,s) and obtained necessary and sufficient conditions to characterize the matrices of classes (ω(p,s),), (ωp(s),) and (ωp(s),)reg.

Matrix transformations between sequence spaces have also been discussed by Savaş and Mursaleen [], Basarir and Savaş [], Mursaleen [, –], Vatan and Simsek [], Savaş [–], Vatan [] and many others.

(2)

In this paper we characterize the matrix classes from this space to the space σ,i.e.,

we obtain necessary and sufficient conditions to characterize the matrices of classes (ω(p,s),

σ), (ωp(s),Vσλ) and (ωp(s),Vσλ)reg.

2 Preliminaries

Letσbe a one-to-one mapping from the setNof natural numbers into itself. A continuous linear functionalϕonl∞is said to be an invariant mean orσ-mean if and only if

(i) ϕ(x)≥when the sequencex= (xk)hasxk≥for allk;

(ii) ϕ(e) = ;

(iii) ϕ(x) =ϕ((k))for allxl∞.

Let denote the set of bounded sequences all of whoseσ-means are equal. We say

that a sequencex= (xk) isσ-convergent if and only ifx. Forσ(n) =n+ , the set

is reduced to the setcˆof almost convergent sequences [, ]. Ifx= (xn), writeTx= ((n)). It is easy to show that

=

xl∞:lim

m tmn(x) =L, uniformly inn;L=σ-limx

,

where

tmn(x) = 

m+  m

j=

Tjxn

andσm(n) denotes themth iterate ofσ atn.

Ifpkis real and positive, we define (see Maddox [])

c(p) =

x: lim

k→∞|xk| pk= 

and

c(p) =x: lim

k→∞|xkl|

pk=  for somel

.

The classes (ω(p,s),), (ωp(s),) and (ωp(s),)reg have been defined by Mohiuddine and Aiyup [] and, forp= (pk) withpk> , the spaceω(p,s) is defined fors≥ by

ω(p,s) =

x: 

n

n

k=

k–s|xkl|pk→,n→ ∞for somel,s≥

,

wheres= (sk) is an arbitrary sequence withsk=  ( k= , , . . .). Ifpk=p, for eachk, we haveω(p,s) =ωp(s).

The sequence space

ω(p) =

x: 

n

n

k=

|xkl|pk→,n→ ∞

(3)

We further define the following.

Letλ= (λm) be a non-decreasing sequence of positive numbers tending to∞such that λm+λm+ , λ= .

A sequencex= (xk) of real numbers is said to be (σ,λ)-convergent to a numberLif and only ifx

σ, where

σ =

xl∞:m→∞lim tmn(x) =L, uniformly inn;L= (σ,λ)-limx

,

tmn(x) =  λm

i∈Im

xσi(n),

andIm= [mλm+ ,m]. Note thatcVσλl∞. Forσ(n) =n+ ,Vσλreduces to the space

ˆ

of almostλ-convergent sequences []; and if we takeσ(n) =n+  andλm=m, then

σ reduces toˆc(see []). Further, if we takeλm=m, thenVσλreduces to.

IfEis a subset ofω, then we writeE+for a generalized Köthe-Toeplitz dual ofE;i.e.,

E+=

a: k

akxkconverges for everyxE

.

If  <pk≤, thenω+(p) =M, where

M=

a: ∞

r=

max

r

r·N–pk|a k|

<∞for some integerN> 

,

and max is the maximum taken over rk< r+(see Theorem , []).

IfXis a topological linear space, we denote byX∗the continuous dual ofX;i.e., the set of all continuous linear functionals onX. Obviously,

ω(p,s)∗=

a: ∞

r=

max

r

r·N–

pkak

sk

<∞for some integerN> 

.

3 Main results

LetXandYbe two nonempty subsets of the spacewof complex sequences. LetA= (ank) (n,k= , , . . .) be an infinite matrix of complex numbers. We writeAx= (An(x)) ifAn(x) :=

kankxkconverges for eachn. (Throughout,

kwill denote summation overkfromk=  tok=∞.) Ifx= (xk)∈Ximplies thatAx= (An(x))∈Y, we say thatAdefines a (matrix) transformation fromXtoY and we denote it byA:XY. By (X,Y) we mean the class of matricesAsuch thatA:XY.

We now characterize the matrices in the class (c(p),Vσλ(p)). We write

tm,n(Ax) =

k

a(n,k,m)xk,

where

a(n,k,m) =  λm

i∈Im

(4)

Theorem . Let <pk≤,then A∈(ω(p,s),Vσλ)if and only if (i) there exists an integerB> such that for everyn

Dn=sup m

r=

max

r

r·B–

pka(n,k,m)

sk <∞;

(ii) α(k)={ank}∞n=Vσλfor eachk; (iii) α={kank}∞n=Vσλ.

In this case theσ-limof Ax is(limx)[ukuk] +kukxkfor every xw(p,s),where

u=σ-lima and uk=σ-lima(k),k= , , . . . .

Proof Suppose thatA∈(ω(p,s),

σ). Defineek= (, , . . . , , , . . .) having  in thekth

coor-dinate sequence. Sinceeandekare inω(p,s), necessity of (ii) and (iii) is clear. We know that

ka(n,k,m)xk converges for eachm,nandxω(p,s). Therefore (a(n,k,m))kω+(p,s) and

r=

max

r

r·B–

pka(n,k,m)

sk <∞

for eachm,n(see []). Furthermore, if fmn(x) =tmn(Ax), then{fmn}m is a sequence of continuous linear functionals onω(p,s) such thatlimmtmn(Ax) exists. Therefore, by using the Banach-Steinhaus theorem, the necessity of (i) follows immediately.

Conversely, suppose that the conditions (i), (ii) and (iii) hold andxω(p,s). We know that (a(n,k,m))k anduk are inω+(p,s) and that the series

ka(n,k,m)xk and

kukxk converge for eachm,n. Write

c(n,k,m) =a(n,k,m) –uk.

Then

k

a(n,k,m)xk=

k

ukxk+l

k

c(n,k,m) + k

c(n,k,m)(xkl)

by (ii) for some integerk> , we have

lim

m

k≤k

c(n,k,m)(xkl) = , uniformly inn,

wherelis the limit ofxforxω(p,s). Since

sup

m,n

r

max

r

r·B–

pkc(n,k,m)Dn,

lim

m

k≤k

a(n,k,m) –uk

sk

sk(xkl)= ,

uniformly inn, whence

lim

n

k

a(n,k,m)xkl·u+

k

(5)

Theorem . Let≤pk<∞,then A∈(ωp(s),Vσλ)if and only if (i) for everyn

M(A) =sup

m

r

rp

r

a(n,skk,m)q

q

<∞,

wherep–+q–= ;

(ii) α(k)∈Vσλfor eachk; (iii) α

σ.

Proof Assume that the conditions are satisfied and letxωp(s). Then tmn(Ax)≤

r=

r

a(n,k,m)skxk

sk

r=

r

a(n,skk,m)q

q

·

r

|xk|p

p

,

and hencetmn(Ax) is absolutely and uniformly convergent for eachm,n. Note that (i) and (ii) imply that

r=

pr

r

|skuk| 

q

M(A) <∞,

so that by Hölder’s inequality,kukxk<∞. Now as in the converse part of Theorem ., it follows thatA∈(ωp(s),Vσλ).

Conversely, suppose thatA∈(ωp(s),Vσλ). Sinceekandeare inωp(s), the necessity of (ii) and (iii) is clear. For the necessity of (i), suppose that

tmn(Ax) =

k

a(n,k,m)xk

exits for eachnwheneverxωp(s). Then, for eachnandr≥, write

fnr(x) =

r

a(n,k,m)xk.

Then{fnr}mis a sequence of continuous linear functionals onωp(s). Since fnr(x)≤

r

a(n,skk,m)q

q

·

r

|sk·xk|p

p

≤pr r

a(n,skk,m)q

q

· x,

it follows that for eachn,

lim

j j

r=

fmr(x) =tmn(Ax)

is in the dual spaceωp∗. Hence there exists aKmnsuch that

(6)

For eachnand any integerj> , definexωp(s) as in [] (Theorem , p.), we get j

r=

pr

r

a(n,skk,m)q

q

Kmn.

Hence, for eachn,

r=

pr

r

a(n,skk,m)q

q

Kmn<∞. (.)

Sincetmn(Ax) is absolutely convergent, we get

tmn(Ax)≤ ∞

r=

pr

r

a(n,skk,m)q

q

x,

so that

Kmn≤ ∞

r=

rp

r

a(n,skk,m)q

q

. (.)

By virtue of (.) and (.),

Kmn= ∞

r=

pr

r

a(n,skk,m)q

q

.

Finally, by (Theorem , [], p.) for everyn, the existence oflimmtmn(Ax) onωp(s) implies that

sup

m

Kmn=sup m

r=

pr

r

a(n,skk,m)q

q

<∞,

which is (i).

Theorem . Let <pk<∞,then A∈(ωp(s),)regif and only if conditions(i), (ii)with σ-lim= and(iii)withσ-lim= +of Theorem.hold.

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

Both authors completed the paper together. Both authors read and approved the final manuscript.

Author details

1Department of Mathematics, Medeniyet University, Istanbul, Turkey.2Department of Mathematics, Istanbul Ticaret University, Uskudar, Istanbul, Turkey.

Acknowledgements

Dedicated to Professor Hari M Srivastava.

We would like to thank the referees for their valuable comments.

(7)

References

1. Banach, S: Theorie des Operations lineaires. Instytut Matematyczny Polskiej Akademi Nauk, Warsaw (1932) 2. Lorentz, GG: A contribution to the theory of divergent sequences. Acta Math.80, 167-190 (1948) 3. King, JP: Almost summable sequences. Proc. Am. Math. Soc.17, 1219-1258 (1966)

4. Schaefer, P: Infinite matrices and invariant means. Proc. Am. Math. Soc.36, 104-110 (1972)

5. Mursaleen, M: Invariant means and some matrix transformations. Tamkang J. Math.10, 181-184 (1979)

6. Mohiuddine, SA, Aiyup, M: Matrix transformations of strongly convergent sequences into. Filomat24(3), 103-109

(2010)

7. Sava¸s, E, Mursaleen, M: Matrix transformations in some sequence spaces. Istanbul Univ. Fen Fak. Mat. Derg.52, 1-5 (1993)

8. Ba¸sarir, M, Sava¸s, E: On matrix transformations of some generalized sequence space. Math. Slovaca45(2), 155-162 (1995)

9. Mursaleen, M: On some new invariant matrix methods of summability. Q. J. Math.34, 77-86 (1983)

10. Mursaleen, M, Jarrah, AM, Mohiuddine, SA: Bounded linear operators for some new matrix transformations. Iran. J. Sci. Technol., Trans. A, Sci.33(A2), 169-177 (2010)

11. Mursaleen, M: On infinite matrices and invariant means. Indian J. Pure Appl. Math.10(4), 457-460 (1979)

12. Mursaleen, M, Gaur, AK, Chishti, TA: On some new sequence spaces of invariant means. Acta Math. Hung.75, 185-190 (1997)

13. Mursaleen, M, Sava¸s, E, Aiyub, M, Mohiuddine, SA: Matrix transformations between the spaces of Cesaro sequences and invariant means. Int. J. Math. Math. Sci.200, Article ID 74319 (2006)

14. Mursaleen, M: Some matrix transformations on sequence spaces of invariant means. Hacet. J. Math. Stat.38(3), 259-264 (2009)

15. Mursaleen, M: OnA-invariant mean andA-almost convergence. Anal. Math.37(3), 173-180 (2011)

16. Mursaleen, M, Mohiuddine, SA: Some matrix transformations of convex and paranormed sequence spaces into the spaces of invariant means. J. Funct. Spaces Appl.2012, Article ID 612671 (2012)

17. Karakaya, V, Simsek, N: On some matrix transformations. Int. Math. J.4(1), 19-25 (2003)

18. Sava¸s, E: On infinite matrices and lacunaryσ-convergence. Appl. Math. Comput.218(3), 1036-1040 (2011) 19. Sava¸s, E: Matrix transformations of some generalized sequence spaces. J. Orissa Math. Soc.4(1), 37-51 (1985) 20. Sava¸s, E: Matrix transformations and absolute almost convergence. Bull. Inst. Math. Acad. Sin.15(3), 345-355 (1987) 21. Sava¸s, E: Matrix transformations between some new sequence spaces. Tamkang J. Math.19(4), 75-80 (1988) 22. Sava¸s, E:σ-summable sequences and matrix transformations. Chin. J. Math.18(3), 201-207 (1990) 23. Sava¸s, E: Matrix transformations and almost convergence. Math. Stud.59(1-4), 170-176 (1991) 24. Sava¸s, E: Matrix transformations ofXpintoCs. Punjab Univ. J. Math.24, 59-66 (1991)

25. Karakaya, V:θσ-summable sequences and some matrix transformations. Tamkang J. Math.35(4), 313-320 (2004)

26. Qamaruddin, Mohiuddine, SA: Almost convergence and some matrix transformations. Filomat21(2), 261-266 (2007) 27. Maddox, IJ: Spaces of strongly summable sequences. Q. J. Math.18, 345-355 (1967)

28. Sava¸s, E: Strongly almost convergence and almostλ-statistical convergence. Hokkaido Math. J.29, 63-68 (2000) 29. Nanda, S: Infinite matrices and almost convergence. J. Indian Math. Soc.40, 173-184 (1976)

30. Maddox, IJ: Elements of Functional Analysis. Cambridge University Press, Cambridge (1970)

31. Lascarides, CG, Maddox, IJ: Matrix transformations between some classes of sequences. Proc. Camb. Philos. Soc.68, 99-104 (1970)

doi:10.1186/1029-242X-2013-254

Cite this article as:Sava¸s and Sava¸s:On some new matrix transformations.Journal of Inequalities and Applications2013

References

Related documents