441 Internat. J. Math. & Math. Sci. Vol. 8 No. 3 (1985) 441-448 NORM-PRESERVING L-L INTEGRAL TRANSFORMATIONS YU CHUEN WEI Department of Mathematics University of Wisconsin-Oshkosh Oshkosh, Wisconsin 54901 U.S.A. 22, (Received March In this paper we consider an L-L integral transformation ABSTRACT. fG(x,y)f(y)dy, F(x) f(y) 1984) is defined on G transformation G(x,y) where [0,). is only a necessary condition, where f01G,(x,t) Idx G,(x,t) fIG,(x,t) Idx =limh+0inf O} O, y for almost all i i of the form and For an L-L integral The following results are proved: to be norm-preserving, G’s. {(x,y): x D is defined on G [t+h t G(x,y)dy t t for each I for almost all t 0 is a necessary and sufficient condition for preserving the norm of certain f L. In this paper 0. x For certain - G the analogous result for sum-preserving L-L integral transformation method. L-L integral transformation. Fubini-Tonelli Theorem. KEY WORDS AND PHRASES. of integrals. 1980 MATHEMATICS SUBJECT CLASSIFICATION CODE. I. into 2. Absolutely continuity - The well-known summability method defined by a matrix A g into E, is sum-preserving if and only if for each k, G present study we also discuss conditions under which L L, is proved. 44A02, 44A06, 42A76. INTRODUCTION. from 0 (ank), mapping i. In>lank G(x,y), defined by In our mappings is norm-preserving or sum-preserving. NOTATION. The notation and terms used are: The statement that U > 0, limit as f if u f is Lebesgue integrable on [0,u] is Lebesgue integrable on . means that for every ff(x)dx L the space of functions that are Lebesgue integrable on D the first quadrant of the plane, i.e., D all G GL L x > 0, where f is defined on C the collection of all the subcollection o[ G [0,) of the form such th,t tends to a finite [0,) {(x,y): x 0, y of the form (*) F(x) F, G- an integral transformation, G: f [or [0,) and that F and G(x,y) with norm 0}. fG(x,y)f(y)dy, defined on D. (*). L whenever f L. and essentially bounded on the space of functions which ar measurable ess fi [0,) with norm s,,Px>Olf(x) 442 3. Y.C. WEI MAIN THEOREM THEOREM i. If G c GL f c L and for every fIF(x) ]dx fIf(y)Idy then for almost all y > 0, fIG,(x,y) lax {y+h lira inf G(x,t)dt, for each x > O. -Y h-O O} satisfying {y Suppose that there is a set A G,(x,y) where PROOF. that either = fIG,(x,y) Idx L, > i for each x for every measurable set A Since G c 1 y c A for all O, it follows from Theorem generality, we can assume that A Case i). y c A, Suppose that for all such for all see [i], y c A. that without loss of is a bounded measurable set. fo[G,(x,y) Idx f0]G,(x,y)]dx c , i < i. Without loss of generality y c A we assume that for each where (T.S.T.), fAG(X,y)dy of finite measure, 0 < mA <fIG,(x,y) Idx or is a small positive number. I < Let c, A(y) f(y) then f G(x,y) A(Y)dY F(x) [A G(x,y)dy and I111- fIf o G(x,y)A(Y)dyldx <-- f0 fA IG(x’y) Idydx Since for each x >_ O, G,(x,y) G(x,y) for almost all y > O, see [2, Theorem 5. P. 255], so it follows from the Fubini-Tonelli Theorem that F[[ <_. f .{A[G(x,y)[dydx A f]G,(x,Y) ldxdy fA(l Hence, for case i), G c)dy is not norm-preserving. > i. Suppose that for all y generality, we assume that for each y e A Case ii). fo[G(x,y) [dx where is a small positive number. e for all x [0,). If F(x) 0 F and + f(y) for almost all XA(y); x F(x) then [0,), AG(X,y)dy then 0 < mA we’ re done. Suppose that G c Let > i Without loss of GL, F(x) 0 for all x in some set with positive measure. so it follows from Theorem (T. S. T) by author, see [I], G,(x,y) Since is measur- 443 NORM-PRESERVING L-L INTEGRAL TRANSFORMATIONS D able on foIG,(x,y) Idx and < M y > O, for almost all M where is a constant. Thus f0"f [G*(x’y) ldydx A A 0 [G*(x’y)]dxdy > O, there is an i > /2 > Given Idydx XO AlG,(x,y) =. < X n < mA/2 O’ for all A y A0, y ..fO Jolc.,(x,y) Idx and from y + o [C,(x,y) [dx for each < e mA/4 having positive measure and for < /8 A y that ]G,(x,y)Idx i / 3/4 {(x,y) [0,X0] 0} AO: IG,(x,y) < /24X E}. Then 0 <mEy < X [0,X0]: (Xl,Y) 0 Let E {x 0. E let i X such that A c A, 0 It follows that there is at least a subset A y satisfying all > 0 0 1 and for any y c for all A0" S ince JA SO ’G,(x,y)Idxdy SAf<=O IG,(x,y> ,dxdy <_ 0 so it follows from the absolute continuity of the integral that there is a x A satisfying mH < 6, and that for every measurable set H 0 6 > 0 such _= [O,xo] ffHiG,(x,y) [dydx < JA If G(x,y)dy 0 0 x > 0, for almost all FII mAo/4 G(x,y) X then A0(Y)dY J’IIAoO-IIAOII G(x,y)dyldx -0 <A and we’re done. By the __= [O,Xo] AO set F and G,(x,y) our on A 0 x E [0,X0] Y of H such that if F. is continuous over uiN__ F [0,X0] 1 the value of (x,y") ’) Generalizati0n0 0 for some set of x >_ 0 with G(x,y)dy of Luzin’s Theorem we can choose a closed [O,Xo] A0 F then mH < 6 and It is clear that G,(x,y) N of subsets A Thus we can have a finite number such that (x,y’); IG,(x, E F. A So we suppose that positive measure. [O,Xo] -G,(x,y") x e [O,X 0], < x A i G,(x,y) X2xio i and within each strip [0,X0] , such that G,(x,y) > O, < O, F 0 (A Ai if (x,y) if (x,y) c F 0 (A Ai) i and ,..,[G,(x,y[ i i E, (x,y") A each there are three sets Then for i and / G,(x,y) mA x A 124X0 if (x,y) e E y > 0 of set for each More precisely, for are close to one another. (x,y’), is uniformly continu- of A i A+i, A and Y.E. WEI 444 (x,y) e F N Hence, if [0,X0] then Ai, IIA. 01 0 G(x,y) A.(y)dyldx G(x,y)dyldx y JA+. f A i G, (x y) dydx + i f f (A A i G,(x,y)dy)dx i Y =JAi fA+ +IAiA(-G,(x,y))dydx G.(x,y)dxdy 1 y f + UA A . i r y l 13.(x y)l dydx i +rE IfAi 13. (x y) dy dx y i Xo fA ]G(x,y) ldydx_ JE SA ]G(x,y) idydx fE IS G, dyi S S’< *<x’y’ I<’>’<’x <S, S,< I,<x.y, <,>, S,< AI,<x.>,,<,yl <,x I. fOiG,(x,Y) idxdy- SE /A.iG,<x,y) idydx I0 i + Y i y >_. y i dx i 2 y 1 (I + 3e/4)mA i If (x ,y) i m{H N [0,X0] 0 AiJ F II-- (slnce mAi/8 e for some A e i N {Ai} I, Y < X 0) then for such an A i, SiS G(x,y) XA i(y)dyldx fol.f. G(x’y)dyldx I,s,,,. o,<x,>,><,>,.<,x > mE flS. + G.(x,y)dyl=i Slio .S, ,ISA G,(x y).yldx (x,y) e H (x,y) e F 0[Ai G,(x,y)dyldx > (i > If m{H such that [0 X 0] Ai} + 3gl4)mA i mAi # 0 XA II, for all IJ emA./8z and we’re done. A H0[0,X0]xA i i e N {Ai} I IG, (x’y) then there is at least an A. dydx < ( mA0/4) mAi/mA0 NORM-PRESERVING L-L INTEGRAL TRANSFORMATION and for such an 445 A i, [,F., =0 [y: A.(y)dyIdx G(x,y) I s=0 G,(x,y) A (y)dyldx + s; s: 1 <o IJAi (x,y) f0 I i G, (x,Y) dY iSAi dx 0 G,(x,y) XA i (y) dy dx G,(,y)dyl d (x,y) : H F G,(x y)dYldx- n. mAi/4 (x,y) e F >_ + 3e/4)mA i (i mAi/8 2 e (I + el2)mA. > mA i :11A G Hence, case (ii) we have proved that norm-preserving and so the proof is is not complete. Theorem I shows us that if G e GL, for almost all y >_ 0, f0iF(x) ldx J0[f(y)lay y > O, J[G,(x,y) [dx S:iF(x)idx flf(Y)Idy for every f e L. is a necessary condition for whenever example will tell us that for almost all cient condition for JoIG,(x,y) Idx f e I e. 1 The next is not a suffi- fiG,(x,y) But the following theorem will show that for certain G’s, [dx =i is a necessary and sufficient condition for preserving the norms of certain f e L. Example. Define -i/4x I/2 if x if x e [i, ), (0,i), G(x,y) for all i/2x 2 and -2/(y + i) 2 [0 I) if y f y e [i =,) f (y) 10/(y Then if (y)[dy 1) 2 1 2/(y + 0 =-2(y + l)2dy + I 101(y + l)2dy 1)-llo =-1+2+5=6 and f0[G,(x,y) ldx f i 0 I/4xll2dx + 1 i/2x 2 dx 2xi/2/4i0 1_ l/2x i<,> i 112 + 112 But F(x) 0 G(x,y) f (y)dy f (-I /4xi/2) 0 0 1 f (y) dy, (i/2x2) f (Y) dY’ if x e (0,i) if x [i, (R)) y > 0; Y.C. WEI 446 where (i/4xi12 0 f(y)dy I =-i/4xl/2[f-2/(y + l)2dy + lO/(y + l)2dy] /4xl/2 -2 (-l) (y + i)-i[0i + lO(-l(y + i)-iii _i =-i/4xi/2[i _l/x 2 + 5] 1/2 g x if (0,i) and (l12x 2)(I 2/x 2 + 5) 2 x e [i ) if Therefore -i + 2(-l)x 2xi/2! 0i =2+2 THEOREM 2. G e GL; Jolf(y) ly 6 4 i) F f whenever f e L and ii) IIF f whenever f e and iii) F f whenever iv) f G,(x,y)dx PROOF it is clear that iv). i) f(y) > 0 f(y) < 0 D and ]0 G(x,y) f(y)dy and JoIF(x)Idx is equivalent to i). G(x,y) > 0 D on O. y F(x) [0,); [O,m); on on 0 G(x,y)If(y) F(x) L, if f If(y) Idy, Assuming that f(y)G(x,y) > 0 have is a nonnegative function on for almost all i, f Since g IIF II= to G(x,y) Suppose that then the folowing are equivalent; on We now prove that and f(y) > 0 i) for all is equivalent y e [0,), Hence D. F(x) G(x,y)f(y)dy _0 0 so IF (x) _ F (x) Therefore and IIF I1: S0IF(x)ldx S F G(x,y)f(y)dydx 0 By the Fubini-Tonelli Theorem and for each all y O, IIF 0 F(x)dx f(Y) 0 x >_ O, C,(x,y) G,(x,y)dxdy Hence F )0 G,(x,y)dx f i if and only if for almost all y _>_ 0 C(x,y) for almost we NORM-PRESERVING L-L INTEGRAL TRANSFORMATIONS Next we prove th 447 Let iii) is equivalent to iv). f(y), if f(y) > 0 0 if f(y) < 0 -f(y), if f(y) < 0 f+ f-0 Since G(x,y) 0 F+ D, on 0 F f (y) > 0 f e L so whenever G(x’Y)f+(y)dy if 0 for all x C(x,y)f (y)dy >_ 0 for all x > 0 0 Z and f(y) / G(x,y) If(y)Idy, F(x) if f+ f- F+ + F- then IF(x) It follows from i) that SO 1F (x) F(x)l flf(Y)IdY if and only if 0 G,(x,y)dx y > 0 for almost all i We are also interested in the analogous sum-preserving question for L-L integral f(y)dy whenever f e L? F(x)dx transformations, viz., when is 0 Next we give the definition of sum-preserving for L-L integral transformations s and aresult concerning it. DEFINITION. GL G The integral transformation is said to be sum-preserving if and only if 0 for all f(y) e L, COROLLARY. then f G where 0 f(y)dy fO G(x,y)f(y)dy. F(x) Suppose that F(x)dx G(x,y) is a nonnegative function on is a sum-preserving transformation whenever G,(x,y)dx i for almost all PROOF. Since L, f y >_ f+ f e L where (y), if f(y) 0 if f (y) < 0 0 f+ -f(y), if f(y) < 0 if f(y) f 0 and IO f(y)dy f] f+dy- f] f-dy and if and only if 0. f-, D >_ 0 G e GL; Y.C. WEI 448 Then I I G(x’Y)[f+I G(x’Y)f+(y)dy- I F(x) G(x,y)f(y)dy f-]dy and 01 6(x’Y)f+(y)dy- F(x)dx 6(x,y)f-(y)dy]dx 0 G(x’Y)f+(y)dydx- J0 G(x,y)f-(y)dy 0 G(x,y)f (y)dydx By the Fubini-Tonelli Theorem 0 G(x’Y)f+(y)dydx (Y) 0 and s f; 0 O(x,y)f (y)dydx f (y) 0 0 s= 0 G,(x,y)dxdy G,(x,y)dxdy Thus O(x,y) f+(y)dydx f+(y)dy 0 and J0 O(x,y) f (y)dydx f (y)dy 0 if and only if 0 G,(x,y)dx i for almost all y 0 Therefore if and only if f7 f 0 if and only if I (::,,:)d 0 I O,(x,y)dx F(x)dx G,(x,y)dx 0 0 f dy for almost all y >_0 f(y)dy 1 for almost all y >_0 The proof is completed. REFERENCES [. WEI, I.C. 2. NATONSON, I.P. Theory of functions of a real variable. Unger Publishing1961. 3. SUNOUCHI, G.[. and TSUCHIKURA, T. Absolute regularity for convergent integrals, Tohoku Math. J. (2) 4 (1952), 153-156. 4. TATCHELL, J.B. On some integral transformation, Proc. London Math. Soc. (3) (1953), 257-266. 5. HARDY, G.H. 6. FRIDY, J.A. Absolute summability matrices that are stronger than the identity mapping, Proc. Amer. Math. Soc. 47 (1975), I|2-118. A property of L-L integral transformation, (will be published). Vol. I, Frederick Divergent series, Clarendon Press, Oxford, 1949.