Selections associated to McShane’s extension theorem for Lipschitz functions

Abstract

Authors

Costica Mustata
“Tiberiu Popoviciu” Institute of Numerical Analysis, Romanian Academy, Romania

Keywords

Paper coordinates

C. Mustăţa, Selections associated to McShane’s extension theorem for Lipschitz functions, Rev. Anal. Numér. Théor. Approx. 21 (1992) 2, 135-145 (MR # 94h: 46008).

PDF

About this paper

Journal

Revue d’Analyse Numer. Theor. Approximation

Publisher Name

Publishing Romanian Academy

Print ISSN

2457-6794

Online ISSN

2501-059X

MR # 94h: 46008

google scholar link

[1] Alfsen, E.M., Effros, E., Structure in real Banach spaces, Ann. of Math. 96 (1972), 98-173.
[2] Azipser, J., Géher, L., Extension of Functions satisfying a Lipschitz conditions Acta Math. Acad. Sci. Hungar 6 (1955), 213-220.
[3] Deutsch, F.., Wu Li, Sung-Ho Park, Tietze Extensions and Conditions Selections for Metric Projecitons J.A.T. 64 (1991), 55-68.
[4] Fakhoury, H., SéIections linéaires associées au théorème de Hahn-Banach, J. Funct. Analysis  11 (7972),436-452.
[5] Mc shane; E, J., Extensions of range of functions, Bull, Amer. Math. Soc. 40 (1934), 837-842.
[6] Mustata C., Best Approximation and Unique Extension of Lipschilz Functions’J.A.T. 19 (1977), 222-230.
[7] Mustata, C., M- ideals in metric spaccs, “Babes-Bolyai,, University, Fac. of Math. and Physics, Research Seminars, Seminar on Math. Anal., preprint Nr. 7 (1988), 65-74.
8. Rudin W., Functional Analysis, McGraw-Hill 1973.

Paper (preprint) in HTML form

1992-Mustata-Selections associated to McShane’s extension-Jnaat

SELECTIONS ASSOCIATED TO McSHANE'S EXTENSION THEOREM FOR LIPSCHITZ FUNCTIONS

COSTICA MUSTATA(Cluj-Napocal)

  1. Let ( X , d X , d X,dX, dX,d ) be a metric space and Y Y YYY a nonvoid subset of X X XXX. A fung tion f : Y R f : Y R f:Y rarr Rf: Y \rightarrow Rf:YR is called Lipschitz on Y ¯ Y ¯ bar(Y)\bar{Y}Y¯ if there exists K 0 K 0 K >= 0K \geqslant 0K0. such that
    ( 1 , 1 1 ) 1 , | f ( x ) f ( y ) | I d ( x , y ) 1 , 1 1 1 , | f ( x ) f ( y ) | I d ( x , y ) (1,(1)/(1))_(1),quad:'|f(x)-f(y)| <= Id(x,y)\left(1, \frac{1}{1}\right)_{1}, \quad \because|f(x)-f(y)| \leqslant I d(x, y)(1,11)1,|f(x)f(y)|Id(x,y),
    for all x , y Y x , y Y x,y in Yx, y \in Yx,yY. A number K 0 K 0 K >= 0K \geqslant 0K0 verifying (1.1) is called a Lipschiti constant for f f fff.
It is easy to show that the quantity f X f X ||f||_(X)\|f\|_{X}fX defined by
(1.2) f X = sup { | f ( x ) f ( y ) | / d ( x , y ) : x , y Y , x y } (1.2) f X = sup { | f ( x ) f ( y ) | / d ( x , y ) : x , y Y , x y } {:(1.2)||f||_(X)=s u p{|f(x)-f(y)|//d(x","y):x","y in Y","x!=y}:}\begin{equation*} \|f\|_{X}=\sup \{|f(x)-f(y)| / \mathrm{d}(x, y): x, y \in Y, x \neq y\} \tag{1.2} \end{equation*}(1.2)fX=sup{|f(x)f(y)|/d(x,y):x,yY,xy}
is the smallest Lipschitz constant for f f fff and we shall call it the Lipschitz norm of f f fff.
Denote by Lip Y Y YYY the set of all real-valued Lipsehitz functions on Y . Y Y_(". ")Y_{\text {. }}Y The set Lip X X XXX and the quantity f x f x ||f||_(x)\|f\|_{x}fx are defined similarly.
Obviously, with respect to the usual operations of addition and multiplication by scalars for functions, Lip Y Y YYY and Lip X X XXX are linear spaces and the functionals Y Y ||||_(Y):}\left\|\|_{Y}\right.Y and X X ||||_(X)\| \|_{X}X are seminorms on these spaces. The functionals r r ||||_(r):}\left\|\|_{r}\right.r and x x ||||_(x)\| \|_{x}x are not norms because they vanish on constant functions.
Mc Shane [5] proved the following extension theorem for Lipschitz functions:
Theorem 1. Let ( X , d ) ( X , d ) (X,d)(X, d)(X,d) be a metric space and Y X Y X O/_(!=Y sub X)\varnothing_{\neq Y \subset X}YX. Then every Lipschitz function on Y Y YYY has a norm preserving extensions to X X XXX, i.e. for every f f f inf \inf Lip Y Y YYY there exists F F F inF \inF Lip X X XXX such that
F | Y = f and F x = f Y . F Y = f  and  F x = f Y . F|_(Y)=f" and "||F||_(x)=||f||_(Y).\left.F\right|_{Y}=f \text { and }\|F\|_{x}=\|f\|_{Y} .F|Y=f and Fx=fY.
Now let x 0 Y x 0 Y x_(0)in Yx_{0} \in Yx0Y be fixed and denote by Lip 0 Y 0 Y _(0)Y{ }_{0} Y0Y the subspace of Lip Y Y YYY formed of all functions in Lip Y Y YYY vanishing at x 0 x 0 x_(0)x_{0}x0, i.e.
(1.3) Lip θ Y = { f Lip Y : f ( x θ ) = 0 } . (1.3) Lip θ Y = f Lip Y : f x θ = 0 . {:(1.3)Lip_(theta)Y={f in Lip Y:f(x_(theta))=0}.:}\begin{equation*} \operatorname{Lip}_{\theta} Y=\left\{f \in \operatorname{Lip} Y: f\left(x_{\theta}\right)=0\right\} . \tag{1.3} \end{equation*}(1.3)LipθY={fLipY:f(xθ)=0}.
s. As F x = f y F x = f y ||F||_(x)=||f||_(y)\|F\|_{x}=\|f\|_{y}Fx=fy, for all F E ( f ) F E ( f ) F in E(f)F \in E(f)FE(f), it follows that E ( f ) E ( f ) E(f)E(f)E(f) is bounded.
To prove the closedness of E ( f ) E ( f ) E(f)E(f)E(f), let ( F n ) F n (F_(n))\left(F_{n}\right)(Fn) be a sequence in E ( f ) E ( f ) E(f)E(f)E(f) converging to a function F Lip p 0 X F Lip p 0 X F^(')in Lipp_(0)XF^{\prime} \in \operatorname{Lip} p_{0} XFLipp0X. As l ( x 0 ) = 0 = F n ( x 0 ) l x 0 = 0 = F n x 0 l^(')(x_(0))=0=F_(n)(x_(0))l^{\prime}\left(x_{0}\right)=0=F_{n}\left(x_{0}\right)l(x0)=0=Fn(x0), for all n N n N n in Nn \in NnN, it follows that
| F n ( x ) : F ( x ) | = | F n ( x ) + F ( x ) ( F n ( x 0 ) F ( x 0 ) ) | F n F x d ( x ; x 0 ) 0 F n ( x ) : F ( x ) = F n ( x ) + F ( x ) F n x 0 F x 0 F n F x d x ; x 0 0 {:[|F_(n)(x):-F(x)|=|F_(n)(x)+F(x)-(F_(n)(x_(0))-F(x_(0)))|cdots <= ],[ <= ||F_(n)-F||_(x)d(x;x_(0))rarr0]:}\begin{aligned} \left|F_{n}(x):-F(x)\right| & =\left|F_{n}(x)+F(x)-\left(F_{n}\left(x_{0}\right)-F\left(x_{0}\right)\right)\right| \cdots \leqslant \\ & \leqslant\left\|F_{n}-F\right\|_{x} \mathrm{~d}\left(x ; x_{0}\right) \rightarrow 0 \end{aligned}|Fn(x):F(x)|=|Fn(x)+F(x)(Fn(x0)F(x0))|FnFx d(x;x0)0
for every x X x X x in Xx \in XxX, implying the pointwise convergence of the sequence ( F n ( x ) ) F n ( x ) (F_(n)(x))\left(F_{n}(x)\right)(Fn(x)) to F ( x ) F ( x ) F^(')(x)F^{\prime}(x)F(x), for all x X x X x in Xx \in XxX. Since F n ( x ) = f ( x ) F n ( x ) = f ( x ) F_(n)(x)=f(x)F_{n}(x)=f(x)Fn(x)=f(x) for all x T x T x inT^(TT)x \in T^{\top}xT and all n N n N n in Nn \in NnN it. follows F ( x ) = f ( x ) F ( x ) = f ( x ) F(x)=f(x)F(x)=f(x)F(x)=f(x), for all x Y x Y x in Yx \in YxY. Also F n F 2 F n F 2 F_(n)rarrF^(2)F_{n} \rightarrow F^{2}FnF2 in Lip 0 X 0 X _(0)X{ }_{0} X0X and F n x = f x , n N F n x = f x , n N ||F_(n)||_(x)=||f||_(x),n in N\left\|F_{n}\right\|_{x}=\|f\|_{x}, n \in NFnx=fx,nN, imply F x = lim n F n x = f y F x = lim n F n x = f y ||F^(')||_(x)=lim_(n rarr oo)||F_(n)||_(x)=||f||_(y)\left\|F^{\prime}\right\|_{x}=\lim _{n \rightarrow \infty}\left\|F_{n}\right\|_{x}=\|f\|_{y}Fx=limnFnx=fy showing that F F ( f ) F F ( f ) F^(')inF(f)F^{\prime} \in \mathbb{F}(f)FF(f);
2 2 2^(@)2^{\circ}2. To prove the second inequality in (1.8) suppose, on the contrary, that there exists x 1 X x 1 X x_(1)in Xx_{1} \in Xx1X such that F ( x 1 ) > F 2 ( x 1 ) F x 1 > F 2 x 1 F^(')(x_(1)) > F_(2)(x_(1))F^{\prime}\left(x_{1}\right)>F_{2}\left(x_{1}\right)F(x1)>F2(x1). Since the functions F F FFF, F 2 F 2 F_(2)F_{2}F2 are continuous on X X XXX and F | X = f = F 2 | Y F X = f = F 2 Y F^(')|_(X)=f=F_(2)|_(Y)\left.F^{\prime}\right|_{X}=f=\left.F_{2}\right|_{Y}F|X=f=F2|Y, it follows that they agree on the closure Y ¯ Y ¯ bar(Y)\bar{Y}Y¯ of Y Y YYY. Therefore x 1 X Y ¯ x 1 X Y ¯ x_(1)in X\\ bar(Y)x_{1} \in X \backslash \bar{Y}x1XY¯, implying d ( x 1 , y ) > 0 d x 1 , y > 0 d(x_(1),y) > 0\mathrm{d}\left(x_{1}, y\right)>0d(x1,y)>0 for all y Y y Y y in Yy \in YyY. Taking into account definition (1.6) of F 2 F 2 F_(2)F_{2}F2 and the inequality F 2 ( x 1 ) < F ( x 1 ) F 2 x 1 < F x 1 F_(2)(x_(1)) < F(x_(1))F_{2}\left(x_{1}\right)< F\left(x_{1}\right)F2(x1)<F(x1), there exists an element y 1 Y y 1 Y y_(1)in Yy_{1} \in Yy1Y such that
(1.9) f ( y 1 ) + f Y d ( x 1 , y 1 ) < F ( x 1 ) . (1.9) f y 1 + f Y d x 1 , y 1 < F x 1 . {:(1.9)f(y_(1))+||f||_(Y)d(x_(1),y_(1)) < F(x_(1)).:}\begin{equation*} f\left(y_{1}\right)+\|f\|_{Y} \mathrm{~d}\left(x_{1}, y_{1}\right)<F\left(x_{1}\right) . \tag{1.9} \end{equation*}(1.9)f(y1)+fY d(x1,y1)<F(x1).
As f ( y 1 ) = F ( y 1 ) f y 1 = F y 1 f(y_(1))=F(y_(1))f\left(y_{1}\right)=F\left(y_{1}\right)f(y1)=F(y1), the inequality (1.9), gives the contradiction f r = f r = ||f||_(r)=\|f\|_{r}=fr= : = F X ( F ( x 1 ) F ( y 1 ) ) / d ( x 1 , y 1 ) > f Y = F X F x 1 F y 1 / d x 1 , y 1 > f Y =||F||_(X) >= (F^(')(x_(1))-F^(')(y_(1)))//d(x_(1),y_(1)) > ||f||_(Y)=\|F\|_{X} \geqslant\left(F^{\prime}\left(x_{1}\right)-F^{\prime}\left(y_{1}\right)\right) / \mathrm{d}\left(x_{1}, y_{1}\right)>\|f\|_{Y}=FX(F(x1)F(y1))/d(x1,y1)>fY.
The first inequality in (1,8) can be proved in a similar way.
3 3 3^(@)3^{\circ}3. Let B Lip 0 Y , B Lip 0 x B Lip 0 Y , B Lip 0 x B_(Lip_(0)Y),B_(Lip_(0)x)B_{\mathrm{Lip}_{0} \mathrm{Y}}, B_{\mathrm{Lip}_{0} \mathrm{x}}BLip0Y,BLip0x be the closed unit balls of the spaces Lip 0 Y 0 Y _(0)Y{ }_{0} Y0Y respectively Lip 0 X Lip 0 X Lip_(0)X\operatorname{Lip}_{0} XLip0X. Let f f fff be an extremal point of B Lipy B Lipy  B_("Lipy ")B_{\text {Lipy }}BLipy  and stippose that H 1 , H 2 B Lip , x H 1 , H 2 B Lip  , x H_(1),H_(2)inB_("Lip ",x)H_{1}, H_{2} \in B_{\text {Lip }, \mathrm{x}}H1,H2BLip ,x and λ ( 0 , 1 ) λ ( 0 , 1 ) lambda in(0,1)\lambda \in(0,1)λ(0,1) are such that λ H 1 + ( 1 λ ) H 2 E ( f ) λ H 1 + ( 1 λ ) H 2 E ( f ) lambdaH_(1)+(1-lambda)H_(2)in E(f)\lambda H_{1}+(1-\lambda) H_{2} \in E(f)λH1+(1λ)H2E(f). Since f f fff is extremal the equality λ H 1 | Y + ( 1 λ ) H 2 | Y = f λ H 1 Y + ( 1 λ ) H 2 Y = f lambdaH_(1)|_(Y)+(1-lambda)H_(2)|_(Y)=f\left.\lambda H_{1}\right|_{Y}+\left.(1-\lambda) H_{2}\right|_{Y}=fλH1|Y+(1λ)H2|Y=f implies H 1 | Y == f = H 2 | x H 1 Y == f = H 2 x H_(1)|_(Y)==f=H_(2)|_(x)\left.H_{1}\right|_{Y}= =f=\left.H_{2}\right|_{x}H1|Y==f=H2|x. Also 1 = f Y λ H 1 + ( 1 λ ) H 2 x λ H x x + ( 1 λ ) H 2 x 1 1 = f Y λ H 1 + ( 1 λ ) H 2 x λ H x x + ( 1 λ ) H 2 x 1 1=||f||_(Y) <= ||lambdaH_(1)+(1-lambda)H_(2)||_(x) <= lambda||H_(x)||_(x)+(1quad lambda)||H_(2)||_(x) <= 11=\|f\|_{Y} \leqslant\left\|\lambda H_{1}+(1-\lambda) H_{2}\right\|_{x} \leqslant \lambda\left\|H_{x}\right\|_{x}+(1 \quad \lambda) \left\|H_{2}\right\|_{x} \leqslant 11=fYλH1+(1λ)H2xλHxx+(1λ)H2x1 implies H 1 x = H 2 x = 1 = f x H 1 x = H 2 x = 1 = f x ||H_(1)||_(x)=||H_(2)||_(x)=1=||f||_(x)\left\|H_{1}\right\|_{x}=\left\|H_{2}\right\|_{x}=1=\|f\|_{x}H1x=H2x=1=fx, showing that H 1 , H 2 E ( f ) H 1 , H 2 E ( f ) H_(1),H_(2)in E(f)H_{1}, H_{2} \in E(f)H1,H2E(f).
Now, suppose that f 2 f Y = 1 f 2 f Y = 1 f_(2)||f||_(Y)=1f_{2}\|f\|_{Y}=1f2fY=1, is not an extremal point of B Lip 0 Y B Lip 0 Y B_(Lip_(0)Y)B_{\mathrm{Lip}_{0} Y}BLip0Y. Then there exist two distinct elements f 1 , f 2 f 1 , f 2 f_(1),f_(2)f_{1}, f_{2}f1,f2 in B Lip 0 Y B Lip  0 Y B_("Lip "_(0))YB_{\text {Lip }_{0}} YBLip 0Y and λ ( 0 , 1 ) λ ( 0 , 1 ) lambda in(0,1)\lambda \in(0,1)λ(0,1) such that f = λ f 1 + ( 1 λ ) f 2 f = λ f 1 + ( 1 λ ) f 2 f=lambdaf_(1)+(1-lambda)f_(2)f=\lambda f_{1}+(1-\lambda) f_{2}f=λf1+(1λ)f2. If H i E ( f i ) , i = 1 , 2 H i E f i , i = 1 , 2 H_(i)in E(f_(i)),i=1,2H_{i} \in E\left(f_{i}\right), i=1,2HiE(fi),i=1,2, then H 1 | x + ( 1 λ ) H 2 | x == λ f 1 + ( 1 λ ) f 2 = f H 1 x + ( 1 λ ) H 2 x == λ f 1 + ( 1 λ ) f 2 = f H_(1)|_(x)+(1-lambda)H_(2)|_(x)==lambdaf_(1)+(1-lambda)f_(2)=f\left.H_{1}\right|_{x}+\left.(1-\lambda) H_{2}\right|_{x}= =\lambda f_{1}+(1-\lambda) f_{2}=fH1|x+(1λ)H2|x==λf1+(1λ)f2=f and 1 = f x = λ H 1 x + ( 1 λ ) H 2 | x x ⩽⩽ λ H 1 + ( 1 λ ) H 2 x 1 1 = f x = λ H 1 x + ( 1 λ ) H 2 x x ⩽⩽ λ H 1 + ( 1 λ ) H 2 x 1 1=||f||_(x)=||lambdaH_(1)||_(x)+(1-lambda)H_(2)|_(x)||_(x)⩽⩽||lambdaH_(1)+(1-lambda)H_(2)||_(x) <= 11=\|f\|_{x}=\left\|\lambda H_{1}\right\|_{x}+\left.(1-\lambda) H_{2}\right|_{x} \|_{x} \leqslant \leqslant\left\|\lambda H_{1}+(1-\lambda) H_{2}\right\|_{x} \leqslant 11=fx=λH1x+(1λ)H2|xx⩽⩽λH1+(1λ)H2x1, showing that λ H 1 + ( 1 λ ) H 2 E ( f ) λ H 1 + ( 1 λ ) H 2 E ( f ) lambdaH_(1)+(1-lambda)H_(2)in E(f)\lambda H_{1}+(1-\lambda) H_{2} \in E(f)λH1+(1λ)H2E(f). Since H i | r = f i f , i = 1 , 2 H i r = f i f , i = 1 , 2 H_(i)|_(r)=f_(i)!=f,i=1,2\left.H_{i}\right|_{r}=f_{i} \neq f, i=1,2Hi|r=fif,i=1,2, it follows that H i E ( f ) , i = 1 , 2 H i E ( f ) , i = 1 , 2 H_(i)in E(f),i=1,2H_{i} \in E(f), i=1,2HiE(f),i=1,2 and E ( f ) E ( f ) E(f)E(f)E(f) is not a face of B p 0 x x B p 0 x x B_(/_\p_(0)x)xB_{\triangle \mathrm{p}_{0} x} xBp0xx.
Remark 1. The extensions F 1 , F 2 F 1 , F 2 F_(1),F_(2)F_{1}, F_{2}F1,F2 of x x xxx function f Lip p 0 Y f Lip p 0 Y f in Lipp_(0)Yf \in \operatorname{Lip} p_{0} YfLipp0Y, given by ( 1.5 ) ( 1.5 ) (1.5)(1.5)(1.5) and ( 1.6 ) ( 1.6 ) (1.6)(1.6)(1.6), are extremal points of the face E ( f ) E ( f ) E(f)E(f)E(f) and consequently they are extremal elements of the unit ball B 1 + D 0 x B 1 + D 0 x B_(1+D_(0)x)B_{1+D_{0} x}B1+D0x, too. Indeed, if F , G ∈∈ B ( f ) F , G ∈∈ B ( f ) F,G∈∈B(f)F, G \in \in B(f)F,G∈∈B(f) and λ ( 0 , 1 ) λ ( 0 , 1 ) lambda in(0,1)\lambda \in(0,1)λ(0,1) are such that F 1 = λ F + ( 1 λ ) G F 1 = λ F + ( 1 λ ) G F_(1)=lambda F+(1-lambda)GF_{1}=\lambda F+(1-\lambda) GF1=λF+(1λ)G then, by (1.8), F 1 F F 1 F F_(1) <= FF_{1} \leqslant FF1F and F 1 G F 1 G F_(1) <= GF_{1} \leqslant GF1G implying F = F 1 = G F = F 1 = G F=F_(1)=GF=F_{1}=GF=F1=G. The extremality of F 2 F 2 F_(2)F_{2}F2 is proved similarly.

2. Let

Y = { F n i Lip 0 X n : F n i M Y j = 0 } , Y = F n i Lip 0 X n : F n i M Y j = 0 , Y^(_|_)={F_(n_(i))inLip_(0)X_(n^(')):F_(n_(i))^(M)∣Y_(j)=0},Y^{\perp}=\left\{F_{n_{i}} \in \operatorname{Lip}_{0} X_{n^{\prime}}: F_{n_{i}}^{M} \mid Y_{j}=0\right\},Y={FniLip0Xn:FniMYj=0},
be the annihilator subspace of Y Y YYY in and X X XXX.
(1.8) F 1 ( x ) F ( x ) F 2 ( x ) , x X , (1.8) F 1 ( x ) F ( x ) F 2 ( x ) , x X , {:(1.8)F_(1)(x) <= F(x) <= F_(2)(x)","x in X",":}\begin{equation*} F_{1}(x) \leqslant F(x) \leqslant F_{2}(x), x \in X, \tag{1.8} \end{equation*}(1.8)F1(x)F(x)F2(x),xX,
where F 1 , F 2 F 1 , F 2 F_(1),F_(2)F_{1}, F_{2}F1,F2 are the extensions of f f fff given by (1.5) and (1.6) ;
3 3 3^(@)3^{\circ}3 For f Y = 1 f Y = 1 ||f||_(Y)=1\|f\|_{Y}=1fY=1, the set E ( f ) E ( f ) E(f)E(f)E(f) is a face of the unit ball B Lip X B Lip  X B_("Lip "_(X))B_{\text {Lip }{ }_{X}}BLip X if and only if f f fff is an extremal point of the unit ball of Lip 0 Y 0 Y _(0)Y{ }_{0} \mathrm{Y}0Y.
Proof. 1 1 1^(@)1^{\circ}1. Let F , G E ( f ) F , G E ( f ) F,G in E(f)F, G \in E(f)F,GE(f) and λ [ 0 , 1 ] λ [ 0 , 1 ] lambda in[0,1]\lambda \in[0,1]λ[0,1]. Obviously that ( λ F + ( 1 λ ) G ) | X = f ( λ F + ( 1 λ ) G ) X = f (lambda F+(1--lambda)G)|_(X)=f(\lambda F+(1- -\lambda) G)\left.\right|_{X}=f(λF+(1λ)G)|X=f and λ F + ( 1 λ ) G X λ F X + ( 1 λ ) G X == λ f X + ( 1 λ ) f Y = f Y λ F + ( 1 λ ) G X λ F X + ( 1 λ ) G X == λ f X + ( 1 λ ) f Y = f Y ||lambda F+(1-lambda)G||_(X) <= lambda||F||_(X)+(1-lambda)||G||_(X)==lambda||f^(')||_(X)+(1-lambda)||f||_(Y)=||f||_(Y)\|\lambda F+(1-\lambda) G\|_{X} \leqslant \lambda\|F\|_{X}+(1-\lambda)\|G\|_{X}= =\lambda\left\|f^{\prime}\right\|_{X}+(1-\lambda)\|f\|_{Y}=\|f\|_{Y}λF+(1λ)GXλFX+(1λ)GX==λfX+(1λ)fY=fY. Since λ F + ( 1 λ ) G X ( λ F + + ( 1 λ ) G | Y Y = f Y λ F + ( 1 λ ) G X ( λ F + + ( 1 λ ) G Y Y = f Y ||lambda F+(1-lambda)G||_(X) >= ||(lambda F++(1-lambda)G|_(Y)||_(Y)=||f||_(Y)\|\lambda F+(1-\lambda) G\|_{X} \geqslant \|(\lambda F+ +\left.(1-\lambda) G\right|_{Y}\left\|_{Y}=\right\| f \|_{Y}λF+(1λ)GX(λF++(1λ)G|YY=fY, it follows that λ F + ( 1 λ ) G E ( f ) λ F + ( 1 λ ) G E ( f ) lambdaF^(')+(1-lambda)G in E(f)\lambda F^{\prime}+(1-\lambda) G \in E(f)λF+(1λ)GE(f), proving the convexity of E ( f ) E ( f ) E(f)E(f)E(f).
A subset Λ Λ Lambda\LambdaΛ of Lip 0 X 0 X _(0)X{ }_{0} X0X is called proximinal in Lip 0 X 0 X _(0)X{ }_{0} X0X if every F Lip 0 X F Lip 0 X F inLip_(0)XF \in \operatorname{Lip}_{0} XFLip0X has a nearest point in Λ Λ Lambda\LambdaΛ, i.e. there exists G A G A G in AG \in AGA such that F G x ¯ =⇋ d ( F , Λ ) F G x ¯ =⇋ d ( F , Λ ) ||F-G||_( bar(x))= ⇋d(F,Lambda)\|F-G\|_{\bar{x}}= \leftrightharpoons \mathrm{d}(F, \Lambda)FGx¯=⇋d(F,Λ), where
d ( F , Λ ) = inf { F H X : H Λ } . d ( F , Λ ) = inf F H X : H Λ . d(F,Lambda)=i n f'{||F-H||_(X):H in Lambda}.\mathrm{d}(F, \Lambda)=\inf ^{\prime}\left\{\|F-H\|_{X}: H \in \Lambda\right\} .d(F,Λ)=inf{FHX:HΛ}.
The metric projection P Λ : P Λ : P_(Lambda):P_{\Lambda}:PΛ: Lip 0 X 2 Λ 0 X 2 Λ _(0)X rarr2^(Lambda){ }_{0} X \rightarrow 2^{\Lambda}0X2Λ is defined by
P Λ ( F ) = { G Λ : F G x = d ( F , A ) } . P Λ ( F ) = G Λ : F G x = d ( F , A ) . P_(Lambda)(F)={G in Lambda:||F-G||_(x)=d(F,(A))}.P_{\Lambda}(F)=\left\{G \in \Lambda:\|F-G\|_{x}=\mathrm{d}(F, \mathrm{~A})\right\} .PΛ(F)={GΛ:FGx=d(F, A)}.
If P Λ ( F ) P Λ ( F ) P_(Lambda)(F)P_{\Lambda}(F)PΛ(F) is a singleton for every F Lip 0 X F Lip 0 X F inLip_(0)XF \in \operatorname{Lip}_{0} XFLip0X then Λ Λ Lambda\LambdaΛ is called a Ohebyshevian subset of: Lip 0 X 0 X _(0)X_{0} X0X.
There is a closed relation between the extension operator E E EEE and the projection pperator P Y P Y P_(Y^(_|_))P_{Y^{\perp}}PY, ilustrated in the following theorem:
Theorem 3. The following assertions hold:
1 1 1^(@)1^{\circ}1 The subspace Y Y Y^(_|_)Y^{\perp}Y is proximinal in Lip 0 X 0 X _(0)X_{0} X0X;
2 2 2^(@)2^{\circ}2 The equality
(2.1) d ( F , Y ) F ) = F H (2.1) d F , Y F = F H {:(2.1){:d(F^('),Y^(_|_))≐||F^('))=||F^(')H^('):}\begin{equation*} \left.\mathrm{d}\left(F^{\prime}, \mathrm{Y}^{\perp}\right) \doteq \| F^{\prime}\right)=\| F^{\prime} H^{\prime} \tag{2.1} \end{equation*}(2.1)d(F,Y)F)=FH
is true for all F Lip 0 X F Lip 0 X F inLip_(0)XF \in \operatorname{Lip}_{0} XFLip0X;
3 3 3^(@)3^{\circ}3 A function G Y G Y G in Y _|_G \in Y \perpGY is a best approximation element for F F FFF if and only if there exists H E ( F | Y ) H E F Y H in E(F|_(Y))H \in E\left(\left.F\right|_{Y}\right)HE(F|Y) such that G = F H G = F H G=F-HG=F-HG=FH, or equivalently
(2.2) P Y ( F ) = F E ( F | Y ) . (2.2) P Y ( F ) = F E F Y . {:(2.2)P_(Y _|_)(F)=F-E(F|_(Y)).:}\begin{equation*} P_{Y \perp}(F)=F-E\left(\left.F\right|_{Y}\right) . \tag{2.2} \end{equation*}(2.2)PY(F)=FE(F|Y).
Proof. First we prove formula (2.1). If F Lip 0 X then for any G Y F Lip 0 X then for any  G Y F inLip_(0)X_("then for any ")G inY^(_|_)F \in \operatorname{Lip}_{0} X_{\text {then for any }} G \in \mathbf{Y}^{\perp}FLip0Xthen for any GY,
F | x Y = sup { | F ( y ) F ( y ) | / d ( y , y ) : y , y Y , y y } = = sup { | ( F G ) ( y ) ( F G ) ( y ) | / d ( y , y ) : y , y Y , y y } sup { | ( F G ) ( x ) ( F G ) ( x ) | / d ( x , x ) : x , x X , x x } = F x Y = sup F ( y ) F y / d y , y : y , y Y , y y = = sup ( F G ) ( y ) ( F G ) y / d y , y : y , y Y , y y sup ( F G ) ( x ) ( F G ) x / d x , x : x , x X , x x = {:[||F|_(x)||_(Y)=s u p{|F^(')(y)-F(y^('))|//d(y,y^(')):y,y^(')in Y,y!=y^(')}=],[=s u p{|(F-G)(y)-(F-G)(y^('))|//d(y,y^(')):y,y^(')in Y,y!=y^(')}],[ <= s u p{|(F-G)(x)-(F-G)(x^('))|//d(x,x^(')):x,x^(')in X,x!=x^(')}],[=]:}\begin{aligned} \left\|\left.F\right|_{x}\right\|_{Y} & =\sup \left\{\left|F^{\prime}(y)-F\left(y^{\prime}\right)\right| / \mathrm{d}\left(y, y^{\prime}\right): y, y^{\prime} \in Y, y \neq y^{\prime}\right\}= \\ & =\sup \left\{\left|(F-G)(y)-(F-G)\left(y^{\prime}\right)\right| / \mathrm{d}\left(y, y^{\prime}\right): y, y^{\prime} \in Y, y \neq y^{\prime}\right\} \\ & \leqslant \sup \left\{\left|(F-G)(x)-(F-G)\left(x^{\prime}\right)\right| / \mathrm{d}\left(x, x^{\prime}\right): x, x^{\prime} \in X, x \neq x^{\prime}\right\} \\ & = \end{aligned}F|xY=sup{|F(y)F(y)|/d(y,y):y,yY,yy}==sup{|(FG)(y)(FG)(y)|/d(y,y):y,yY,yy}sup{|(FG)(x)(FG)(x)|/d(x,x):x,xX,xx}=
implying that F | Y Y d ( F , Y ) F Y Y d F , Y ||F|_(Y)||_(Y) <= d(F,Y^(_|_))\left\|\left.F\right|_{Y}\right\|_{Y} \leqslant \mathrm{~d}\left(F, Y^{\perp}\right)F|YY d(F,Y).
On the other hand, by Theorem 1, there exists H H H inH \inH Lip 0 X 0 X _(0)X{ }_{0} X0X such that H | Y = F | Y H Y = F Y H|_(Y)=F|_(Y)\left.H\right|_{Y}=\left.F\right|_{Y}H|Y=F|Y and H X = F Y | Y Y H X = F Y Y Y ||H||_(X)=||F^(Y)|_(Y)||_(Y)\|H\|_{X}=\left\|\left.F^{Y}\right|_{Y}\right\|_{Y}HX=FY|YY. It follows that F F FFF. H Y H Y H inY^(_|_)H \in Y^{\perp}HY and : F Y | Y Y = F ( F H ) x inf { F G x : G Y } = d ( F , Y ) F Y Y Y = F ( F H ) x inf F G x : G Y = d F , Y ||F^(Y)|_(Y)||_(Y)=||F-(F-H)_(x)|| >= i n f{||F-G||_(x):G inY^(_|_)}=d(F^('),Y _|_)\left\|\left.F^{Y}\right|_{Y}\right\|_{Y} =\left\|F-(F-H)_{x}\right\| \geqslant \inf \left\{\|F-G\|_{x}: G \in Y^{\perp}\right\}=\mathrm{d}\left(F^{\prime}, Y \perp\right)FY|YY=F(FH)xinf{FGx:GY}=d(F,Y), showing that F | Y Y = d ( F , Y ) F Y Y = d F , Y ||F|_(Y)||_(Y)=d(F,Y^(_|_))\left\|\left.F\right|_{Y}\right\|_{Y}=\mathrm{d}\left(F, Y^{\perp}\right)F|YY=d(F,Y).
Assertion 3 3 3^(@)3^{\circ}3 and formula (2.2) follow from [6], Lemma 1, p. 223 and assertion 1 1 1^(@)1^{\circ}1 follows from 3 3 3^(@)3^{\circ}3.
Now, by Theorem 2, 1 1 1^(@)1^{\circ}1, the set P Y ( T ) = F E ( F | Y ) P Y ( T ) = F E F Y P_(Y)(T)=F-E(F|_(Y))P_{Y}(T)=F-E\left(\left.F\right|_{Y}\right)PY(T)=FE(F|Y) is bounded, convex and closed, for any F Lip 0 X F Lip 0 X F inLip_(0)XF \in \operatorname{Lip}_{0} XFLip0X.
We shall say that the set Y X Y X Y sub XY \subset XYX has the property ( U ) ( U ) (U)(U)(U) if every f ∈∈ Lip 0 Y f ∈∈ Lip 0 Y f∈∈Lip_(0)Yf \in \in \operatorname{Lip}_{0} Yf∈∈Lip0Y has a unique Lipschitz extension F Lip 0 X F Lip 0 X F^(')inLip_(0)XF^{\prime} \in \operatorname{Lip}_{0} XFLip0X, i.e. E ( f ) E ( f ) E(f)E(f)E(f) is a singleton for every f Lip p 0 Y f Lip p 0 Y f in Lipp_(0)Yf \in \operatorname{Lip} p_{0} YfLipp0Y. By Theorem 3, the set Y Y YYY has property ( U ) ( U ) (U)(U)(U) if and only if Y Y Y^(_|_)Y^{\perp}Y is a Chebyshevian subspace of Lip 0 X Lip 0 X Lip_(0)X\operatorname{Lip}_{0} XLip0X.
2. A natural question is when have the set valued operators E E EEE and P Y P Y P_(Y _|_)P_{Y \perp}PY continuous selections. If S : A 2 B S : A 2 B S:A rarr2^(B)S: A \rightarrow 2^{\mathrm{B}}S:A2B is a set-valued application, a function s : A B s : A B s:A rarr Bs: A \rightarrow Bs:AB is called a selection for S S SSS if s ( x ) S ( x ) s ( x ) S ( x ) s(x)in S(x)s(x) \in S(x)s(x)S(x), for all x A x A x in Ax \in AxA.
In the following theorems, we shall prove the existence of continuous selections for the operators E E EEE and P Y . L P Y . L P_(Y.L)P_{Y . L}PY.L in the particular case X = R X = R X=RX=RX=R, with the usual distance d ( x , y ) = | x y | d ( x , y ) = | x y | d(x,y)=|x-y|\mathrm{d}(x, y)=|x-y|d(x,y)=|xy|.
Theorem 4. Let X = R , Y = [ a , b ] R X = R , Y = [ a , b ] R X=R,Y=[a,b]in RX=R, Y=[a, b] \in RX=R,Y=[a,b]R and x 0 [ a , b ] x 0 [ a , b ] x_(0)in[a,b]x_{0} \in[a, b]x0[a,b] fixed. Then the extension operator E : Lip 0 Y 2 Lip 0 R E : Lip 0 Y 2 Lip 0 R E:Lip_(0)Y rarr2^(Lip_(0)R)E: \operatorname{Lip}_{0} Y \rightarrow 2^{\operatorname{Lip}_{0} R}E:Lip0Y2Lip0R has a continuous and positively homogeneous selection e e eee.
Proof. Define ε 2 : Lip 0 Y Lip 0 R ˙ ε 2 : Lip 0 Y Lip 0 R ˙ epsi_(2):Lip_(0)Y rarrLip_(0)R^(˙)\varepsilon_{2}: \operatorname{Lip}_{0} Y \rightarrow \operatorname{Lip}_{0} \dot{R}ε2:Lip0YLip0R˙ by
e 2 ( f ) = F 2 , f Lip 0 Y e 2 ( f ) = F 2 , f Lip 0 Y e_(2)(f)=F_(2),quad f inLip_(0)Ye_{2}(f)=F_{2}, \quad f \in \operatorname{Lip}_{0} Ye2(f)=F2,fLip0Y
where F 2 F 2 F_(2)F_{2}F2 is the maximal extension of f f fff given by (1.6).
If α > 0 α > 0 alpha > 0\alpha>0α>0 then
e 2 ( a f ) ( x ) = in { a f ( y ) + a f Y | x y | : y [ a , b ] } = a : inf { f ( y ) + f Y | x y | 1 : y [ a b ] } = = a F 2 ( x ) e 2 ( a f ) ( x ) = in a f ( y ) + a f Y | x y | : y [ a , b ] = a : inf f ( y ) + f Y | x y | 1 : y [ a b ] = = a F 2 ( x ) {:[e_(2)(af)(x)=in{af(y)+||af||_(Y)|x-y|:y in[a,b]}],[=a:i n f{f(y)+||f||_(Y)|x-y|_(1):y in[a*b]}=],[=a*F_(2)(x)]:}\begin{aligned} e_{2}(a f)(x) & =\operatorname{in}\left\{a f(y)+\|a f\|_{Y}|x-y|: y \in[a, b]\right\} \\ & =a: \inf \left\{f(y)+\|f\|_{Y}|x-y|_{1}: y \in[a \cdot b]\right\}= \\ & =a \cdot F_{2}(x) \end{aligned}e2(af)(x)=in{af(y)+afY|xy|:y[a,b]}=a:inf{f(y)+fY|xy|1:y[ab]}==aF2(x)
showing that e 2 e 2 e_(2)e_{2}e2 is positively homogeneous.
To prove the continuity of e 2 e 2 e_(2)e_{2}e2 for ε > 0 ε > 0 epsi > 0\varepsilon>0ε>0, take δ = ε / 3 δ = ε / 3 delta=epsi//3\delta=\varepsilon / 3δ=ε/3 and show that
(2.3) | e 2 ( f ) e 2 ( g ) | ε , (2.3) e 2 ( f ) e 2 ( g ) ε , {:(2.3)|e_(2)(f)-e_(2)(g)| <= epsi",":}\begin{equation*} \left|e_{2}(f)-e_{2}(g)\right| \leqslant \varepsilon, \tag{2.3} \end{equation*}(2.3)|e2(f)e2(g)|ε,
for all f , g Lip 0 Y f , g Lip 0 Y f,g inLip_(0)Yf, g \in \operatorname{Lip}_{0} Yf,gLip0Y such that f g Y < δ f g Y < δ ||f-g||_(Y) < delta\|f-g\|_{Y}<\deltafgY<δ.
It is easy to check that the maximal extension F 2 F 2 F_(2)F_{2}F2 of f f fff has the form
F 2 ( x ) = f ( a ) f Y ( x a ) , for x < a = f ( x ) , for x [ a , b ] = f ( b ) + f Y ( x b ) , for x > b . F 2 ( x ) = f ( a ) f Y ( x a ) ,  for  x < a = f ( x ) ,  for  x [ a , b ] = f ( b ) + f Y ( x b ) ,  for  x > b . {:[F_(2)(x)=f(a)-||f||_(Y)(x-a)","" for "x < a],[=f(x)","quad" for "x in[a","b]],[=f(b)+||f||_(Y)(x-b)","" for "x > b.]:}\begin{aligned} F_{2}(x) & =f(a)-\|f\|_{Y}(x-a), \text { for } x<a \\ & =f(x), \quad \text { for } x \in[a, b] \\ & =f(b)+\|f\|_{Y}(x-b), \text { for } x>b . \end{aligned}F2(x)=f(a)fY(xa), for x<a=f(x), for x[a,b]=f(b)+fY(xb), for x>b.
If g Lip 0 Y g Lip 0 Y g inLip_(0)Yg \in \operatorname{Lip}_{0} YgLip0Y is such that f g Y < δ = ε / 3 f g Y < δ = ε / 3 ||f-g||_(Y) < delta=epsi//3\|f-g\|_{Y}<\delta=\varepsilon / 3fgY<δ=ε/3, then similarly the maximal extension G 2 G 2 G_(2)G_{2}G2 of g g ggg has the expression:
θ 2 ( x ) = g ( a ) g Y ( x a ) , for x < a = g ( x ) , for x [ a , b ] = g ( b ) + g Y ( x b ) , for x > b . θ 2 ( x ) = g ( a ) g Y ( x a ) ,  for  x < a = g ( x ) ,  for  x [ a , b ] = g ( b ) + g Y ( x b ) ,  for  x > b . {:[theta_(2)(x)=g(a)-||g||_(Y)(x-a)","" for "x < a],[=g(x)","quad" for "x in[a","b]],[=g(b)+||g||_(Y)(x-b)","" for "x > b.]:}\begin{aligned} \theta_{2}(x) & =g(a)-\|g\|_{Y}(x-a), \text { for } x<a \\ & =g(x), \quad \text { for } x \in[a, b] \\ & =g(b)+\|g\|_{Y}(x-b), \text { for } x>b . \end{aligned}θ2(x)=g(a)gY(xa), for x<a=g(x), for x[a,b]=g(b)+gY(xb), for x>b.
It follows that the difference H 2 = F 2 G 2 H 2 = F 2 G 2 H_(2)=F_(2)-G_(2)H_{2}=F_{2}-G_{2}H2=F2G2 has the expression :
H 2 ( x ) = f ( a ) g ( a ) ( f Y g Y ) ( x a ) , for x < a = f ( x ) g ( x ) , = f ( b ) g ( b ) + ( f Y g Y ) ( x b ) , for x > b H 2 ( x ) = f ( a ) g ( a ) f Y g Y ( x a ) ,  for  x < a = f ( x ) g ( x ) , = f ( b ) g ( b ) + f Y g Y ( x b ) ,  for  x > b {:[H_(2)(x)=f(a)-g(a)-(||f||_(Y)-||g||_(Y))(x-a)","" for "x < a],[=f(x)-g(x)","],[=f(b)-g(b)+(||f||_(Y)-||g||_(Y))(x-b)","" for "x > b]:}\begin{aligned} H_{2}(x) & =f(a)-g(a)-\left(\|f\|_{Y}-\|g\|_{Y}\right)(x-a), \text { for } x<a \\ & =f(x)-g(x), \\ & =f(b)-g(b)+\left(\|f\|_{Y}-\|g\|_{Y}\right)(x-b), \text { for } x>b \end{aligned}H2(x)=f(a)g(a)(fYgY)(xa), for x<a=f(x)g(x),=f(b)g(b)+(fYgY)(xb), for x>b
We have to consider several cases:
(... Case 1 x , y > b , x y 1 x , y > b , x y 1^(@):.x,y > b,x!=y1^{\circ} \therefore x, y>b, x \neq y1x,y>b,xy. In this case
(2.4) | H 2 ( x ) H 2 ( y ) | | f X g Y | | x y | f g Y | x y | < ε | x y | (2.4) H 2 ( x ) H 2 ( y ) f X g Y | x y | f g Y | x y | < ε | x y | {:[(2.4)|H_(2)(x)-H_(2)(y)| <= |||f||_(X)-||g||_(Y)|*|x-y|],[ <= ||f-g||_(Y)*|x-y| < epsi|x-y|]:}\begin{align*} \left|H_{2}(x)-H_{2}(y)\right| & \leqslant\left|\|f\|_{X}-\|g\|_{Y}\right| \cdot|x-y| \tag{2.4}\\ & \leqslant\|f-g\|_{Y} \cdot|x-y|<\varepsilon|x-y| \end{align*}(2.4)|H2(x)H2(y)||fXgY||xy|fgY|xy|<ε|xy|
Case 2 . x , y < a , x y 2 . x , y < a , x y 2^(@).x,y < a,x!=y2^{\circ} . x, y<a, x \neq y2.x,y<a,xy. Reasoning like in Case 1 1 1^(@)1^{\circ}1 one obtains
(2.5) | H 2 ( x ) H 2 ( y ) | < ε | x y | . (2.5) H 2 ( x ) H 2 ( y ) < ε | x y | . {:(2.5)|H_(2)(x)-H_(2)(y)| < epsi*|x-y|.:}\begin{equation*} \left|H_{2}(x)-H_{2}(y)\right|<\varepsilon \cdot|x-y| . \tag{2.5} \end{equation*}(2.5)|H2(x)H2(y)|<ε|xy|.
Case 3 : a x , y b ; x y 3 : a x , y b ; x y 3^(@):a <= x,y <= b;x!=y3^{\circ}: a \leqslant x, y \leqslant b ; x \neq y3:ax,yb;xy. In this case
(2.6) | H 2 ( x ) H 2 ( y ) | = | ( f g ) ( x ) ( f g ) ( y ) | | f g | | Y | x y | ( ε / 3 ) | x y | . (2.6) H 2 ( x ) H 2 ( y ) = | ( f g ) ( x ) ( f g ) ( y ) | | f g | Y | x y | ( ε / 3 ) | x y | . {:[(2.6)|H_(2)(x)-H_(2)(y)|=|(f-g)(x)-(f-g)(y)| <= ],[ <= |f-g||_(Y)*|x-y| <= (epsi//3)*|x-y|.]:}\begin{align*} \left|H_{2}(x)-H_{2}(y)\right| & =|(f-g)(x)-(f-g)(y)| \leqslant \tag{2.6}\\ & \leqslant\left.|f-g|\right|_{Y} \cdot|x-y| \leqslant(\varepsilon / 3) \cdot|x-y| . \end{align*}(2.6)|H2(x)H2(y)|=|(fg)(x)(fg)(y)||fg||Y|xy|(ε/3)|xy|.
(2.7) | H 2 ( x ) H 2 ( y ) | =∣ ( f g ) ( x ) ( f g ) ( b ) (2.7) H 2 ( x ) H 2 ( y ) =∣ ( f g ) ( x ) ( f g ) ( b ) {:(2.7)|H_(2)(x)-H_(2)(y)|=∣(f-g)(x)-(f-g)(b)-:}\begin{equation*} \left|H_{2}(x)-H_{2}(y)\right|=\mid(f-g)(x)-(f-g)(b)- \tag{2.7} \end{equation*}(2.7)|H2(x)H2(y)|=∣(fg)(x)(fg)(b)
abor
(1.5) , 11 (1.5)  ,  11 {:(1.5)" , "11:}\begin{equation*} \text { , } 11 \tag{1.5} \end{equation*}(1.5) , 11
Case 5 . y < a 1 x b 5 . y < a 1 x b 5^(@).y < a_(1) <= x <= b5^{\circ} . y<a_{1} \leqslant x \leqslant b5.y<a1xb. Beasoning like in Case 4 4 4^(@)4^{\circ}4 one obtains
(2.8) | H 2 ( x ) H 2 ( y ) | ( 2 ϵ / 3 ) | x y | . (2.8) H 2 ( x ) H 2 ( y ) ( 2 ϵ / 3 ) | x y | . {:(2.8)|H_(2)(x)-H_(2)(y)| <= (2epsilon//3)*|x-y|.:}\begin{equation*} \left|H_{2}(x)-H_{2}(y)\right| \leqslant(2 \epsilon / 3) \cdot|x-y| . \tag{2.8} \end{equation*}(2.8)|H2(x)H2(y)|(2ϵ/3)|xy|.
Case 6 < a < b < y 6 < a < b < y 6^(@)cdots oo < a < b < y6^{\circ} \cdots \infty<a<b<y6<a<b<y. In this case
(2.9) | H 2 ( x ) H 2 ( y ) | =∣ ( f g ) ( a ) ( 1 | f Y g | Y ) ( x ) a ) ( f g ˙ ) ( b ) ( f x g y ) ( y b ) 3 f g Y | a b | + | f Y g Y | | x + y a b | < < 3 | | f g Y | x y | < 3 δ : | x y | = ε | x y | . (2.9) H 2 ( x ) H 2 ( y ) =∣ ( f g ) ( a ) 1 f Y g Y ( x ) a ( f g ˙ ) ( b ) f x g y ( y b ) 3 f g Y | a b | + f Y g Y | x + y a b | < < 3 | | f g Y | x y | < 3 δ : | x y | = ε | x y | . {:[(2.9){:|H_(2)(x)-H_(2)(y)|=∣(f-g)(a)-(1|f||_(Y)-||g|Y)(x)-a)-||],[-(f-g^(˙))(b)-(||f||_(x)-||g||_(y))(y-b)||^(3) <= ],[ <= ||f-g*||_(Y)|a-b|+|||f||_(Y)-||g||_(Y)|*|x+y-a-b| < ],[ < 3*||f-g||_(Y)|x-y| < 3delta:|x-y|=epsi*|x-y|.]:}\begin{align*} & \left.\left|H_{2}(x)-H_{2}(y)\right|=\mid(f-g)(a)-\left(1\left|f\left\|_{Y}-\right\| g\right| Y\right)(x)-a\right)-\| \tag{2.9}\\ & -(f-\dot{g})(b)-\left(\|f\|_{x}-\|g\|_{y}\right)(y-b) \|^{3} \leqslant \\ & \leqslant\|f-g \cdot\|_{Y}|a-b|+\left|\|f\|_{Y}-\|g\|_{Y}\right| \cdot|x+y-a-b|< \\ & <3 \cdot| | f-g \|_{Y}|x-y|<3 \delta:|x-y|=\varepsilon \cdot|x-y| . \end{align*}(2.9)|H2(x)H2(y)|=∣(fg)(a)(1|fYg|Y)(x)a)(fg˙)(b)(fxgy)(yb)3fgY|ab|+|fYgY||x+yab|<<3||fgY|xy|<3δ:|xy|=ε|xy|.
ast Taking into account the inequalities (2.4)-(2.9), it follows that F 2 G 2 X = H 2 X = sup { | H 2 ( x ) H 2 ( y ) | | | x y ∣: x , y R , x y } < ε F 2 G 2 X = H 2 X = sup H 2 ( x ) H 2 ( y ) | | x y ∣: x , y R , x y < ε ||F_(2)-G_(2)||_(X)=||H_(2)||_(X)=s u p{|H_(2)(x)-H_(2)(y)|||x-y∣:x,y in R,x!=y} < epsi\left\|F_{2}-G_{2}\right\|_{X}=\left\|H_{2}\right\|_{X}=\sup \left\{\left|H_{2}(x)-H_{2}(y)\right|| | x-y \mid: x, y \in R, x \neq y\right\}<\varepsilonF2G2X=H2X=sup{|H2(x)H2(y)|||xy∣:x,yR,xy}<ε i.e.
| e 2 ( f ) e 2 ( g ) | < ε . e 2 ( f ) e 2 ( g ) < ε . |e_(2)(f)-e_(2)(g)| < epsi.\left|e_{2}(f)-e_{2}(g)\right|<\varepsilon .|e2(f)e2(g)|<ε.
Remark 2. The selection e 1 ( f ) = F 1 e 1 ( f ) = F 1 e_(1)(f)=F_(1)e_{1}(f)=F_{1}e1(f)=F1, where F 1 F 1 F_(1)F_{1}F1 is the minimal extension of f f fff defined by (1.5) is also continuous and positively homogeneous. This can be proved directly or taking into account the equality
(2.10) e 1 ( f ) = e 2 ( f ) , (2.10) e 1 ( f ) = e 2 ( f ) , {:(2.10)e_(1)(f)=-e_(2)(-f)",":}\begin{equation*} e_{1}(f)=-e_{2}(-f), \tag{2.10} \end{equation*}(2.10)e1(f)=e2(f),
which holds for all f Lip 0 Y f Lip 0 Y f in Lip_(0)Yf \in \operatorname{Lip}{ }_{0} YfLip0Y.
Combining these two results one obtains the following consequence:
Corollary 5. The extension operator E - Lip 0 Y 2 Lip 0 E - Lip  0 Y 2 Lip  0 E_("- Lip "0)Y rarr2^("Lip "0)E_{\text {- Lip } 0} Y \rightarrow 2^{\text {Lip } 0}E- Lip 0Y2Lip 0 R has a contimious and homogeneous selection given by
e ( f ) = ( 1 / 2 ) ( e 1 ( f ) + e 2 ( f ) ) , f Lip 0 Y . e ( f ) = ( 1 / 2 ) e 1 ( f ) + e 2 ( f ) , f Lip 0 Y . e(f)=(1//2)*(e_(1)(f)+e_(2)(f)),f inLip_(0)Y.e(f)=(1 / 2) \cdot\left(e_{1}(f)+e_{2}(f)\right), f \in \operatorname{Lip}_{0} Y .e(f)=(1/2)(e1(f)+e2(f)),fLip0Y.
Proof. Obviously that e e eee is continuous and positively homogeneous. On the other hand by (2.10), it follows
e ( f ) e ( f ) , f Lip 0 Y , e ( f ) e ( f ) , f Lip 0 Y , e(-f)≐-e(f),f inLip_(0)Y,e(-f) \doteq-e(f), f \in \operatorname{Lip}_{0} Y,e(f)e(f),fLip0Y,
implying the homogeneity of e : e ( α f ) α : e ( f ) , α R , f Lip ] 0 I e : e ( α f ) α : e ( f ) , α R , f Lip ] 0 I e:e(alpha f)⇌alpha:e(f),alpha in R,f in Lip]_(0)Ie: e(\alpha f) \rightleftharpoons \alpha: e(f), \alpha \in R, f \in \operatorname{Lip}]_{0} \mathcal{I}e:e(αf)α:e(f),αR,fLip]0I.
3. This section is concerned with the existence of selections for the metric projection P Y P Y P_(Y _|_)P_{Y \perp}PY in the case X R , Y = [ a , b ] , x 0 [ a , b ] X R , Y = [ a , b ] , x 0 [ a , b ] X=>R,Y=[a,b],x_(0)in[a,b]X \Rightarrow R, Y=[a, b], x_{0} \in[a, b]XR,Y=[a,b],x0[a,b]. : usal selection p p ppp : Lip X Y X Y X rarrY^(_|_)X \rightarrow Y^{\perp}XY of P Y P Y P_(Y_(_|_))P_{Y_{\perp}}PY is called additive modulo Y Y Y^(_|_)Y^{\perp}Y, provided:
(81) p ( F + G ) = p ( F ) + p ( G ) , (81) p ( F + G ) = p ( F ) + p ( G ) , {:(81)p(F+G)=p(F)+p(G)",":}\begin{equation*} p(F+G)=p(F)+p(G), \tag{81} \end{equation*}(81)p(F+G)=p(F)+p(G),
formall e 1 p 0 X e 1 p 0 X e^(-1)p_(0)Xe^{-1} p_{0} Xe1p0X and G Y G Y G inY^(_|_)G \in Y^{\perp}GY.
Theorem 6. The metric projection P Y P Y P_(Y _|_)P_{Y \perp}PY has a homogeneous, additive modulo Y Y Y _|_Y \perpY and continuous setection p p ppp.
Proof. Let e e eee be the homogeneous and continuous selection of the extension operator E E EEE, given in Corollary 5. Taking into account Theorem 3 and equality (2.10), define p : Lip 0 X Y p : Lip 0 X Y p:Lip_(0)X rarr Y _|_p: \operatorname{Lip}_{0} X \rightarrow Y \perpp:Lip0XY by the formula
p = I e r , p = I e r , p=I-e@r,p=I-e \circ r,p=Ier,
where r : Lip 0 X Lip 0 Y r : Lip 0 X Lip 0 Y r:Lip_(0)X rarrLip_(0)Yr: \operatorname{Lip}_{0} X \rightarrow \operatorname{Lip}_{0} Yr:Lip0XLip0Y is the restriction operator given by r ( F ) == F | Y r ( F ) == F Y r(F)==F^(')|_(Y)r(F)= =\left.F^{\prime}\right|_{Y}r(F)==F|Y and I : Lip 0 X : Lip 0 X I : Lip 0 X : Lip 0 X I:Lip_(0)X:Lip_(0)XI: \operatorname{Lip}_{0} X: \operatorname{Lip}_{0} XI:Lip0X:Lip0X this the identity map. Then
p ( F ) = ( I e r ) ( F ) = F e ( F | Y ) P Y ( F ) . p ( F ) = ( I e r ) ( F ) = F e F Y P Y ( F ) . p(F)=(I-e@r)(F)=F-e(F|_(Y))inP_(Y _|_)(F).p(F)=(I-e \circ r)(F)=F-e\left(\left.F\right|_{Y}\right) \in P_{Y \perp}(F) .p(F)=(Ier)(F)=Fe(F|Y)PY(F).
Indeed F e ( F | Y ) = ( 1 / 2 ) ( F F 1 ) + ( 1 / 2 ) ( F F 2 ) P Y ( F ) F e F Y = ( 1 / 2 ) F F 1 + ( 1 / 2 ) F F 2 P Y ( F ) F-e(F|_(Y))=(1//2)(F-F_(1))+(1//2)(F-F_(2))inP_(Y _|_)(F)F-e\left(\left.F\right|_{Y}\right)=(1 / 2)\left(F-F_{1}\right)+(1 / 2)\left(F-F_{2}\right) \in P_{Y \perp}(F)Fe(F|Y)=(1/2)(FF1)+(1/2)(FF2)PY(F); since the set P Y ( F ) P Y F P_(Y _|_)(F^('))P_{Y \perp}\left(F^{\prime}\right)PY(F) is convex.
Obviously the selection p p ppp is continuous and
p ( α F ) = α F + e ( α F | Y ) = α ( F + e ( | F | Y ) ) = α p ( F ) ) p ( α F ) = α F + e α F Y = α F + e F Y = α p ( F ) {:p(alpha F)=alpha F+e( alpha F|_(Y))=alpha(F+e(|F^(')|_(Y)))=alpha*p(F)^('))\left.p(\alpha F)=\alpha F+e\left(\left.\alpha F\right|_{Y}\right)=\alpha\left(F+e\left(\left|F^{\prime}\right|_{Y}\right)\right)=\alpha \cdot p(F)^{\prime}\right)p(αF)=αF+e(αF|Y)=α(F+e(|F|Y))=αp(F))
for all α R α R alpha inR\alpha \in \mathbb{R}αR, showing that p p ppp is homogeneous.
Now
(3.2) p ( F + G ) = F + G e ( ( F + G ) | X ) = F + G e ( F | Y ) = = F e ( F | Y ) + G = p ( F ) + G = p ( F ) + p ( G ) (3.2) p ( F + G ) = F + G e ( F + G ) X = F + G e F Y = = F e F Y + G = p ( F ) + G = p ( F ) + p ( G ) {:[(3.2)p(F+G)=F+G-e((F+G)|_(X))=F+G-e(F|_(Y))=],[=F-e(F|_(Y))+G=p(F)+G=p(F)+p(G)]:}\begin{align*} p(F+G) & =F+G-e\left(\left.(F+G)\right|_{X}\right)=F+G-e\left(\left.F\right|_{Y}\right)= \tag{3.2}\\ & =F-e\left(\left.F\right|_{Y}\right)+G=p(F)+G=p(F)+p(G) \end{align*}(3.2)p(F+G)=F+Ge((F+G)|X)=F+Ge(F|Y)==Fe(F|Y)+G=p(F)+G=p(F)+p(G)
for all T Lip p 0 X T Lip p 0 X T in Lipp_(0)XT \in \operatorname{Lip} p_{0} XTLipp0X and G Y G Y G inY^(_|_)G \in Y^{\perp}GY, since for G Y , P Y ( G ) = { G } G Y , P Y ( G ) = { G } G inY^(_|_),P_(Y^(_|_))(G)={G}G \in Y^{\perp}, P_{Y^{\perp}}(G)=\{G\}GY,PY(G)={G} and p ( G ) = G p ( G ) = G p(G)=Gp(G)=Gp(G)=G
By (3.2), the selection p p ppp is additive modulo Y Y Y _|_Y \perpY and Theorem 6 is. proved.
It is easily seen that the kernel of P Y L P Y L P_(YL)P_{Y L}PYL,
Ker P Y = { F Lip 0 X , 0 P Y ( F ) } Ker P Y = F Lip 0 X , 0 P Y ( F ) KerP_(Y _|_)={F inLip_(0)X,0inP_(Y _|_)(F)}\operatorname{Ker} P_{Y \perp}=\left\{F \in \operatorname{Lip}_{0} X, 0 \in P_{Y \perp}(F)\right\}KerPY={FLip0X,0PY(F)}
verifies the equality
(3.3) Ker P Y = { F Lip p 0 X : F X = F | Y Y } (3.3) Ker P Y = F Lip p 0 X : F X = F Y Y {:(3.3)KerP_(Y^('))={F in Lipp_(0)X:||F||_(X)=||F|_(Y)||_(Y)}:}\begin{equation*} \operatorname{Ker} P_{Y^{\prime}}=\left\{F \in \operatorname{Lip} p_{0} X:\|F\|_{X}=\left\|\left.F\right|_{Y}\right\|_{Y}\right\} \tag{3.3} \end{equation*}(3.3)KerPY={FLipp0X:FX=F|YY}
Cozollary 7. For X = R , Y = [ a ; b ] X = R , Y = [ a ; b ] X=R,Y=[a;b]X=R, \mathrm{Y}=[a ; b]X=R,Y=[a;b] and x 0 [ a ; b ] x 0 [ a ; b ] x_(0)in[a;b]x_{0} \in[a ; b]x0[a;b] the following. assertions are true:
(a) The extension operator Thes a linear and continuous setection;
(b) The metric projection P Y P Y P_(Y _|_)P_{Y \perp}PY has linear and continuous selection;
(c) There exists a subspace W W WWW of the subspace Ker P Y P Y P_(Y _|_)P_{Y \perp}PY such that every T Lip 0 X T Lip 0 X T^(')inLip_(0)XT^{\prime} \in \mathrm{Lip}_{0} XTLip0X can be uniquely represented in the form F = H + G F = H + G F=H+GF=H+GF=H+G, with H W , G Y H W , G Y H inW^(TT),G inY^(_|_)H \in W^{\top}, G \in Y^{\perp}HW,GY, i.e. the subspace Y Y Y^(_|_)Y^{\perp}Y is complemented in Lip 0 X 0 X _(0)X{ }_{0} X0X.
Proof. (a) We show that the application e : Lip p 0 Ψ e : Lip p 0 Ψ e:Lipp_(0)Psi rarre: \operatorname{Lip} p_{0} \Psi \rightarrowe:Lipp0Ψ Lip p 0 X defi- p 0 X defi-  p_(0)X_("defi- ")p_{0} X_{\text {defi- }}p0Xdefi-  ned by
(3.4) e ( f ) = ( 1 / 2 ) ( F 1 + F 2 ) , (3.4) e ( f ) = ( 1 / 2 ) F 1 + F 2 , {:(3.4)e(f)=(1//2)(F_(1)+F_(2))",":}\begin{equation*} e(f)=(1 / 2)\left(F_{1}+F_{2}\right), \tag{3.4} \end{equation*}(3.4)e(f)=(1/2)(F1+F2),
where H 1 , F 2 H 1 , F 2 H_(1),F_(2)H_{1}, F_{2}H1,F2 are the extremal extensions of f f fff given by ( 1 , 5 ) ( 1 , 5 ) (1,5)(1,5)(1,5) and ( 1 , 6 ) ( 1 , 6 ) (1,6)(1,6)(1,6), is linear and continuous selection of E E EEE.
Writing explicitly e e eee we find that
e ( f ) ( x ) = f ( a ) , for x < a = f ( x ) , for x [ a , b ] = f ( b ) , for x > b e ( f ) ( x ) = f ( a ) ,  for  x < a = f ( x ) ,  for  x [ a , b ] = f ( b ) ,  for  x > b {:[e(f)(x)=f(a)","" for "x < a],[=f(x)","" for "x in[a","b]],[=f(b)","" for "x > b]:}\begin{aligned} e(f)(x) & =f(a), \text { for } x<a \\ & =f(x), \text { for } x \in[a, b] \\ & =f(b), \text { for } x>b \end{aligned}e(f)(x)=f(a), for x<a=f(x), for x[a,b]=f(b), for x>b
for any f Lip 0 Y f Lip 0 Y f inLip_(0)Yf \in \operatorname{Lip}_{0} \mathbf{Y}fLip0Y. Obviously e ( α f + β g ) ( x ) = α e ( f ) ( x ) + β e ( g ) ( x ) e ( α f + β g ) ( x ) = α e ( f ) ( x ) + β e ( g ) ( x ) e(alpha f+beta g)(x)=alpha*e(f)(x)+beta*e(g)(x)e(\alpha f+\beta g)(x)=\alpha \cdot e(f)(x)+\beta \cdot e(g)(x)e(αf+βg)(x)=αe(f)(x)+βe(g)(x) for all x R x R x in Rx \in RxR and all f , g Lip θ θ Y , α , β R f , g Lip θ θ Y , α , β R f,g in Liptheta_(theta)Y,alpha,beta in Rf, g \in \operatorname{Lip} \theta_{\theta} Y, \alpha, \beta \in Rf,gLipθθY,α,βR, showing that e e eee is linear. The continuity of eiwas proved in Corollary 5.
(b) The application p : Lip 0 Y Y p : Lip 0 Y Y p:Lip_(0)Y rarrY^(_|_)p: \operatorname{Lip}_{0} Y \rightarrow Y^{\perp}p:Lip0YY defined, for F Lip p 0 X F Lip p 0 X F in Lipp_(0)XF \in \operatorname{Lip} p_{0} XFLipp0X, by
(3.5) p ( F ) = F e ( F | Y ) = F ( 1 / 2 ) ( F 1 + F 2 ) (3.5) p ( F ) = F e F Y = F ( 1 / 2 ) F 1 + F 2 {:(3.5)p(F)=F-e(F^(')|_(Y))=F-(1//2)(F_(1)+F_(2)):}\begin{equation*} p(F)=F-e\left(\left.F^{\prime}\right|_{Y}\right)=F-(1 / 2)\left(F_{1}+F_{2}\right) \tag{3.5} \end{equation*}(3.5)p(F)=Fe(F|Y)=F(1/2)(F1+F2)
where F 1 , F 2 F 1 , F 2 F_(1),F_(2)F_{1}, F_{2}F1,F2 are the extensions given by (1.5), (1.6), is linear and continuous.
(c) Let
(3:6) W = { e ( F | X ) : F Lip 0 X } (3:6) W = e F X : F Lip 0 X {:(3:6)W={e(F|_(X)):F inLip_(0)X}:}\begin{equation*} W=\left\{e\left(\left.F\right|_{X}\right): F \in \operatorname{Lip}_{0} X\right\} \tag{3:6} \end{equation*}(3:6)W={e(F|X):FLip0X}
The linearity of e e eee implies that W W WWW is a subspace of Lip p 0 X Lip p 0 X Lipp_(0)X\operatorname{Lip} p_{0} XLipp0X. By (3.3) and the equality
e ( F ) Y X = F | Y Y e ( F ) Y X = F Y Y ||e(F∣)_(Y)||_(X)=||F|_(Y)||_(Y)\left\|e(F \mid)_{Y}\right\|_{X}=\left\|\left.F\right|_{Y}\right\|_{Y}e(F)YX=F|YY
it follows that W Ker P Y W Ker P Y W sub KerP_(Y _|_)W \subset \operatorname{Ker} P_{Y \perp}WKerPY.
For F Lip p 0 X F Lip p 0 X F in Lipp_(0)XF \in \operatorname{Lip} p_{0} XFLipp0X define
(3.7) G ( x ) = 0 , for x [ a , b ] = F ( x ) F ( a ) , for x < a = F ( x ) F ( b ) , for x > b (3.7) G ( x ) = 0 ,  for  x [ a , b ] = F ( x ) F ( a ) ,  for  x < a = F ( x ) F ( b ) ,  for  x > b {:[(3.7)G(x)=0","" for "x in[a","b]],[=F^(')(x)-F(a)","" for "x < a],[=F^(')(x)-F(b)","" for "x > b]:}\begin{align*} G(x) & =0, \text { for } x \in[a, b] \tag{3.7}\\ & =F^{\prime}(x)-F(a), \text { for } x<a \\ & =F^{\prime}(x)-F(b), \text { for } x>b \end{align*}(3.7)G(x)=0, for x[a,b]=F(x)F(a), for x<a=F(x)F(b), for x>b
and
(3.8) H ( x ) = F ¯ ( x ) , for x [ a , b ] = F ( a ) , for x < a = F ( b ) , for x > b (3.8) H ( x ) = F ¯ ( x ) ,  for  x [ a , b ] = F ( a ) ,  for  x < a = F ( b ) ,  for  x > b {:[(3.8)H(x)= bar(F)(x)","" for "x in[a","b]],[=F(a)","" for "x < a],[=F(b)","" for "x > b]:}\begin{align*} H(x) & =\bar{F}(x), \text { for } x \in[a, b] \tag{3.8}\\ & =F(a), \text { for } x<a \\ & =F(b), \text { for } x>b \end{align*}(3.8)H(x)=F¯(x), for x[a,b]=F(a), for x<a=F(b), for x>b
Then F = H + G , G Y F = H + G , G Y F=H+G,G inY^(_|_)F=H+G, G \in Y^{\perp}F=H+G,GY and H = e ( F | Y ) W H = e F Y W H=e(F|_(Y))in WH=e\left(\left.F\right|_{Y}\right) \in WH=e(F|Y)W.
To prove that Lip 0 X Lip 0 X Lip_(0)X\operatorname{Lip}_{0} XLip0X is the topological direct sum of the subspaces Y Y Y^(_|_)Y^{\perp}Y and W W WWW, it is sufficient to prove that the projection operator on X ¯ X ¯ bar(X)^(_|_)\bar{X}^{\perp}X¯ is continuous, i.e. that the application F G F G F rarr GF \rightarrow GFG, where G Y G Y G in Y _|_G \in Y \perpGY is the function defined in (3.7) is a linear and continuous operator from Lip 0 X Lip 0 X Lip_(0)X\operatorname{Lip}_{0} XLip0X to Y Y Y^(_|_)Y^{\perp}Y.
寻 1 The linearity is obvious. To prove the continuity suppose that F n , T 0 , F F n , T 0 , F F_(n),T_(0),FF_{n}, T_{0}, FFn,T0,F in Lip 0 X Lip 0 X Lip_(0)X\operatorname{Lip}_{0} XLip0X, i.e. F n F x 0 F n F x 0 ||F_(n)-F||_(x)rarr0\left\|F_{n}-F\right\|_{x} \rightarrow 0FnFx0. Then
G n ( x ) = 0 , x [ a , b ] = F n ( x ) F n ( a ) , x < a = F n ( x ) F n ( b ) , x > b G n ( x ) = 0 , x [ a , b ] = F n ( x ) F n ( a ) , x < a = F n ( x ) F n ( b ) , x > b {:[G_(n)(x)=0","x in[a","b]],[=F_(n)(x)-F_(n)(a)","x < a],[=F_(n)(x)-F_(n)(b)","x > b]:}\begin{aligned} G_{n}(x) & =0, x \in[a, b] \\ & =F_{n}(x)-F_{n}(a), x<a \\ & =F_{n}(x)-F_{n}(b), x>b \end{aligned}Gn(x)=0,x[a,b]=Fn(x)Fn(a),x<a=Fn(x)Fn(b),x>b
and
U n ( x ) = G n ( x ) G ( x ) = 0 , x [ a ; b ] = F n ( x ) F ( x ) + ( F ( a ) F n ( a ) ) , x < a = F n ( x ) F ( x ) + ( F ( b ) F n ( b ) ) , x > b U n ( x ) = G n ( x ) G ( x ) = 0 , x [ a ; b ] = F n ( x ) F ( x ) + F ( a ) F n ( a ) , x < a = F n ( x ) F ( x ) + F ( b ) F n ( b ) , x > b {:[U_(n)(x)=G_(n)(x)-G(x)=0","quad x in[a;b]],[=F_(n)(x)-F(x)+(F(a)-F_(n)(a))","quad x < a],[=F_(n)(x)-F(x)+(F(b)-F_(n)(b))","quad x > b]:}\begin{aligned} U_{n}(x) & =G_{n}(x)-G(x)=0, \quad x \in[a ; b] \\ & =F_{n}(x)-F(x)+\left(F(a)-F_{n}(a)\right), \quad x<a \\ & =F_{n}(x)-F(x)+\left(F(b)-F_{n}(b)\right), \quad x>b \end{aligned}Un(x)=Gn(x)G(x)=0,x[a;b]=Fn(x)F(x)+(F(a)Fn(a)),x<a=Fn(x)F(x)+(F(b)Fn(b)),x>b
Again, we have to consider "several cases:
Case 1. x , y < a x , y < a x,y < ax, y<ax,y<a. In this case
| U n ( x ) U n ( y ) | = | ( F n F ) ( x ) ( F n F ) ( y ) | F n F X | x y | U n ( x ) U n ( y ) = F n F ( x ) F n F ( y ) F n F X | x y | |U_(n)(x)-U_(n)(y)|=|(F_(n)-F)^(')(x)-(F_(n)-F)(y)| <= ||F_(n)-F||_(X)*|x-y|\left|U_{n}(x)-U_{n}(y)\right|=\left|\left(F_{n}-F\right)^{\prime}(x)-\left(F_{n}-F\right)(y)\right| \leqslant\left\|F_{n}-F\right\|_{X} \cdot|x-y||Un(x)Un(y)|=|(FnF)(x)(FnF)(y)|FnFX|xy|
The same inequality is obtained for x , y > b x , y > b x,y > bx, y>bx,y>b.
Case 2. a < x < b < y a < x < b < y a < x < b < ya<x<b<ya<x<b<y. In this case
| U n ( x ) U n ( y ) | = | U n ( y ) | = | F n ( y ) F ( y ) ( F ( b ) F n ( b ) ) | = = | ( F n F ) ( y ) ( F n F ) ( b ) | F n F x | y b | | | F n F x | x y | U n ( x ) U n ( y ) = U n ( y ) = F n ( y ) F ( y ) F ( b ) F n ( b ) = = F n F ( y ) F n F ( b ) F n F x | y b | F n F x x y {:[|U_(n)(x)-U_(n)(y)|=|U_(n)(y)|=|F_(n)(y)-F(y)-(F(b)-F_(n)(b))|=],[=|(F_(n)-F)(y)-(F_(n)-F)(b)| <= ||F_(n)-F||_(x)|y-b| <= ||F_(n)-F||_(x)*|x-y|]:}\begin{gathered} \left|U_{n}(x)-U_{n}(y)\right|=\left|U_{n}(y)\right|=\left|F_{n}(y)-F(y)-\left(F(b)-F_{n}(b)\right)\right|= \\ =\left|\left(F_{n}-F\right)(y)-\left(F_{n}-F\right)(b)\right| \leqslant\left\|F_{n}-F\right\|_{x}|y-b| \leqslant\left|\left|F_{n}-F \|_{x} \cdot\right| x-y\right| \end{gathered}|Un(x)Un(y)|=|Un(y)|=|Fn(y)F(y)(F(b)Fn(b))|==|(FnF)(y)(FnF)(b)|FnFx|yb|||FnFx|xy|
The same inequality holds for x < a < y ˙ < b x < a < y ˙ < b x < a < y^(˙) < bx<a<\dot{y}<bx<a<y˙<b.
Case 3. x < a < b < y x < a < b < y x < a < b < yx<a<b<yx<a<b<y. In this case
| U n ( x ) U n ( y ) | =∣ F n ( x ) F ( x ) ( F n ( a ) F ( a ) ) F n ( y ) + F ( y ) + + F n ( b ) F ( b ) | | ( F n F ) ( x ) ( F n F ) ( y ) | + | ( F n F ) ( b ) ( F n F ) ( a ) | F n F X | x y + F n F X ( b a ) 2 F n F X | x y | U n ( x ) U n ( y ) =∣ F n ( x ) F ( x ) F n ( a ) F ( a ) F n ( y ) + F ( y ) + + F n ( b ) F ( b ) F n F ( x ) F n F ( y ) + F n F ( b ) F n F ( a ) F n F X x y + F n F X ( b a ) 2 F n F X | x y | {:[|U_(n)(x)-U_(n)(y)|=∣F_(n)(x)-F(x)-(F_(n)(a)-F(a))-F_(n)(y)+F(y)+],[+F_(n)(b)-F(b)| <= |(F_(n)-F)(x)-(F_(n)-F)(y)|+|(F_(n)-F)(b)-],[-(F_(n)-F)(a)| <= ||F_(n)-F||_(X)*|x-y∣+||F_(n)-F||_(X)*(b-a)],[ <= 2||F_(n)-F||_(X)*|x-y|]:}\begin{gathered} \left|U_{n}(x)-U_{n}(y)\right|=\mid F_{n}(x)-F(x)-\left(F_{n}(a)-F(a)\right)-F_{n}(y)+F(y)+ \\ +F_{n}(b)-F(b)\left|\leqslant\left|\left(F_{n}-F\right)(x)-\left(F_{n}-F\right)(y)\right|+\right|\left(F_{n}-F\right)(b)- \\ -\left(F_{n}-F\right)(a)\left|\leqslant\left\|F_{n}-F\right\|_{X} \cdot\right| x-y \mid+\left\|F_{n}-F\right\|_{X} \cdot(b-a) \\ \leqslant 2\left\|F_{n}-F\right\|_{X} \cdot|x-y| \end{gathered}|Un(x)Un(y)|=∣Fn(x)F(x)(Fn(a)F(a))Fn(y)+F(y)++Fn(b)F(b)||(FnF)(x)(FnF)(y)|+|(FnF)(b)(FnF)(a)|FnFX|xy+FnFX(ba)2FnFX|xy|
It follows that
| U n ( x ) U n ( y ) | 2 F n F X | x y | , U n ( x ) U n ( y ) 2 F n F X | x y | , |U_(n)(x)-U_(n)(y)| <= 2||F_(n)-F||_(X)*|x-y|,\left|U_{n}(x)-U_{n}(y)\right| \leqslant 2\left\|F_{n}-F\right\|_{X} \cdot|x-y|,|Un(x)Un(y)|2FnFX|xy|,
for all x , y R x , y R x,y in Rx, y \in Rx,yR, implying D n X 2 F n F X 0 D n X 2 F n F X 0 ||D_(n)||_(X) <= 2||F_(n)-F||_(X)rarr0\left\|D_{n}\right\|_{X} \leqslant 2\left\|F_{n}-F\right\|_{X} \rightarrow 0DnX2FnFX0.
It follows that F n F F n F F_(n)rarr FF_{n} \rightarrow FFnF implies G n G G n G G_(n)rarr GG_{n} \rightarrow GGnG, showing that the projection operator on Y Y Y^(_|_)Y^{\perp}Y is continuous and consequently L 1 p 0 X L 1 p 0 X L^(1)p_(0)XL^{1} p_{0} XL1p0X is the direct sum of Y i Y i Y^(i)Y^{i}Yi and W W WWW. Corollary 7 is completely proved.
: : Remarks 3. (a) In the considered case ( X = R , Y ˙ = [ a , b ] , x 0 [ a , b ] ) X = R , Y ˙ = [ a , b ] , x 0 [ a , b ] ) (X=R,(Y^(˙))=[a,b],x_(0)in:}in[a,b])\left(X=R, \dot{Y}=[a, b], x_{0} \in\right. \in[a, b])(X=R,Y˙=[a,b],x0[a,b]), we have e 1 ( f ) e 2 ( f ) e 1 ( f ) e 2 ( f ) e_(1)(f)!=e_(2)(f)e_{1}(f) \neq e_{2}(f)e1(f)e2(f), for all f Lip 0 Y , f 0 f Lip 0 Y , f 0 f inLip_(0)Y,f!=0f \in \operatorname{Lip}_{0} Y, f \neq 0fLip0Y,f0. Tn fact e 1 ( f ) < e ( f ) << e 2 ( f ) , f Lip 0 Y e 1 ( f ) < e ( f ) << e 2 ( f ) , f Lip 0 Y e_(1)(f) < e(f)<<e_(2)(f),fLip_(0)Ye_{1}(f)<e(f)< <e_{2}(f), f \operatorname{Lip}_{0} Ye1(f)<e(f)<<e2(f),fLip0Y.
(b) Let X = [ 0 , 1 ] X = [ 0 , 1 ] X=[0,1]X=[0,1]X=[0,1] with d ( x , y ) = | x y | , Y = { 0 , 1 } d ( x , y ) = | x y | , Y = { 0 , 1 } d(x,y)=|x-y|,Y={0,1}\mathrm{d}(x, y)=|x-y|, Y=\{0,1\}d(x,y)=|xy|,Y={0,1} and x 0 = 0 x 0 = 0 x_(0)=0x_{0}=0x0=0 Then e 1 ( f ) = e 2 ( f ) = e ( f ) e 1 ( f ) = e 2 ( f ) = e ( f ) e_(1)(f)=e_(2)(f)=e(f)e_{1}(f)=e_{2}(f)=e(f)e1(f)=e2(f)=e(f), for all f Lip 0 Y f Lip 0 Y f inLip_(0)Yf \in \operatorname{Lip}_{0} YfLip0Y. It follows that Y = { H Lip 0 X : F ( 0 ) = F ( 1 ) = 0 } Y = H Lip 0 X : F ( 0 ) = F ( 1 ) = 0 } Y^(_|_)={H^(')inLip_(0)X:}:F(0)=F(1)=0}Y^{\perp}=\left\{H^{\prime} \in \operatorname{Lip}_{0} X\right. : F(0)=F(1)=0\}Y={HLip0X:F(0)=F(1)=0} is a Chebyshevian subspace of Lip 0 X 0 X _(0)X{ }_{0} X0X. In this case Lip 0 X = Ker P Y Y Lip 0 X = Ker P Y Y Lip_(0)X=KerP_(Y _|_)o+Y^(_|_)\operatorname{Lip}_{0} X=\operatorname{Ker} P_{Y \perp} \oplus Y^{\perp}Lip0X=KerPYY and the extension operator E E EEE and the metric projection P Y P Y P_(Y^(_|_))P_{Y^{\perp}}PY are linear and single valued.

REFERENCES

  1. Alfsen, E. M., Effross, E., Stutucture in real Banach spaces, Ann. of Math. 96 (1972),98-173. 2. Czipser, J., Gêher, L., Extension of Functions sotisfying a Lipschitz.conditions. Acta Math. Acad. Sci. Hungạr 6 (1955), 213-220.
  2. Deutsch, F., Wu Li, Sung-Ho Park, Tiedze Explensions and Continuous Selections for Metric Projections. J.A.T. 64 (1991), 55-68.
  3. Fakhoury, H., Sélections linéaires associées au théorème de Hahn-Banach, J. Funct. Analysis 11 (1972), 436-452.
  4. Mc Shane, E. J., Extension of range of functions, Bull. Amer. Math. Soc. 40 (1934), 837-842.
  5. Mustăta, C., Best Approximation and Unique Extension of Lipschitz Functions, J.A.T. 19 (1977), 222-230.
  6. Mustăja, C., M-ideals in metric spaces, "Babeş-Bolyai" University, Fac. of Math. and Physics, Research Seminars, Seminar on Math. Anal., Preprint Nr. 7, 1988, 65-74. 8. Rudin W., Functional Analysis, McGraw-Hill 1973.
Received 15.V. 1992
Institutul de Calcul
Oficiul Poștal 1 C.P. 68
3400 Cluj-Napoca Romania
1992

Related Posts