• 検索結果がありません。

On the critical case of Okamoto's continuous non-differentiable functions

N/A
N/A
Protected

Academic year: 2022

シェア "On the critical case of Okamoto's continuous non-differentiable functions"

Copied!
5
0
0

読み込み中.... (全文を見る)

全文

(1)

On the critical case of Okamoto's continuous non‑differentiable functions

著者 Kobayashi Kenta

journal or

publication title

Proceedings of the Japan Academy Series A:

Mathematical Sciences

volume 85

number 8

page range 101‑104

year 2009‑01‑01

URL http://hdl.handle.net/2297/37862

doi: 10.3792/pjaa.85.101

(2)

On the critical case of Okamoto’s continuous non-dierentiable functions

By Kenta KOBAYASHI

Faculty of Mathematics and Physics, Institute of Science and Engineering, Kanazawa University, Kakuma-machi, Kanazawa, Ishikawa 920-1192, Japan

(Communicated by Masaki KASHIWARA,M.J.A., Sept. 14, 2009)

Abstract: In a recent paper in this Proceedings, H. Okamoto presented a parameterized family of continuous functions which contains Bourbaki’s and Perkins’s nowhere dierentiable functions as well as the Cantor-Lebesgue singular function. He showed that the function changes it’s dierentiability from ‘dierentiable almost everywhere’ to ‘non-dierentiable almost every- where’ at a certain parameter value. However, dierentiability of the function at the critical parameter value remained unknown. For this problem, we prove that the function is non- dierentiable almost everywhere at the critical case.

Key words: Continuous non-dierentiable function; the law of the iterated logarithm.

1. Introduction. We consider a parameter- ized family of continuous functions which were pre- sented by H. Okamoto [3, 4]. This function can be regarded as a generalization of Bourbaki’s [1] and Perkins’s [5] nowhere dierentiable functions as well as of the Cantor-Lebesgue singular function.

Okamoto’s function is constructed as the limit of a sequence ffng1n¼0 of piecewise linear and contin- uous functions. For a xed parametera2 ð0;1Þ, each function in the sequence is dened as follows:

(i) f0ðxÞ ¼x;

(ii) fnþ1ðxÞis continuous on [0,1], (iii) fnþ1 k

3n ¼fn k 3n ; fnþ1 3kþ1

3nþ1

¼fn

k 3n þa fn kþ1

3n

fn k 3n

; fnþ1 3kþ2

3nþ1

¼fn

k 3n þ ð1aÞ fn

kþ1 3n

fn

k 3n

; fnþ1 kþ1

3n

¼fn

kþ1 3n

;

fork¼0;1; ;3n1;

(iv) fnþ1ðxÞis linear in each subinterval k

3nþ1xkþ1 3nþ1 for

k¼0;1; ;3nþ11:

Figure 1 shows the operation from fn to fnþ1. Oka- moto’s functionFaðxÞis then dened as

FaðxÞ ¼ lim

n!1fnðxÞ:

He noticed thatFaðxÞis continuous on½0;1and co- incides with some known functions whenatakes par- ticular values. For example, the cases a¼5=6 and a¼2=3 correspond to nowhere-dierentiable func- tions dened by Perkins [5] and Bourbaki [1] respec- tively. Also, if a¼1=2, Fa is the Cantor-Lebesgue singular function which is non-decreasing and has zero derivative almost everywhere (Fig. 2).

2. Dierentiability of Fa. In the paper [3], H. Okamoto proved thatFaðxÞhas the following fea- tures:

(i) If a < a0, then FaðxÞ is dierentiable almost everywhere.

(ii) If a0< a <2=3, thenFaðxÞis non-dierentiable almost everywhere.

(iii) If 2=3a <1, then FaðxÞis nowhere dieren- tiable.

Here, the constanta0ð¼0:5592 Þis the unique real root of

54a327a2¼1:

2000 Mathematics Subject Classication. Primary 26A27;

Secondary 26A30.

doi: 10.3792/pjaa.85.101 62009 The Japan Academy

(3)

As for the case a¼a0, it remained open whether FaðxÞ is dierentiable almost everywhere or non- dierentiable almost everywhere. In this case, we proved thatFaðxÞis non-dierentiable almost every- where.

3. Main result. The main result of this arti- cle is the following

Theorem 1. If a¼a0, then FaðxÞ is non- dierentiable almost everywhere in½0;1Þ.

In order to prove this theorem, we need some denitions and a preliminary lemma concerning with the law of the iterated logarithm [2].

Denitions. Let x¼X1

n¼1

nðxÞ

3n ; nðxÞ 2 f0;1;2g;

denote the ternary expansion of x2 ½0;1Þ. If x is a rational number of the formk=3n, we use the ternary expansion ending in all 0’s (instead of the one ending in all 2’s). We also use the following notations:

cðkÞ ¼ 1; ðk¼0 or k¼2Þ;

2; ðk¼1Þ;

and

SnðxÞ ¼Xn

k¼1

c kðxÞ

; TnðxÞ ¼1

nþ1ðxÞ ¼1

SnðxÞ;

where 1ðAÞ is the indicator function that takes the value one if argumentAis true and zero otherwise.

With these denitions, we have the following lemma:

Lemma 1.

lim sup

n!1

TnðxÞ ffiffiffin p 1 holds for almost every x2 ½0;1Þ.

Proof.Since thecðnÞare i.i.d. random variables with mean 0 and variance 2 with respect to Lebesgue measure on ð0;1Þ, the law of the iterated logarithm [2] implies that

Fig. 1. The operation fromfntofnþ1. Before the operation (top) and after the operation (bottom). This operation is performed in each subinterval½k=3n;ðkþ1Þ=3n.

Fig. 2. The graph of Perkins’s function (top), Bourbaki’s function (middle) and the Cantor-Lebesgue singular function (bottom).

(4)

lim sup

n!1

SnðxÞ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 4nlog logn p ¼1 and

lim inf

n!1

SnðxÞ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 4nlog logn p ¼ 1 almost everywhere inð0;1Þ.

Thus in particular, the events SnðxÞ=pffiffiffin 1 and SnðxÞ=pffiffiffin

1 both happen innitely often.

Each time SnðxÞ=pffiffiffin

exits the interval ½1;1Þ, it must do so with a value of

c

nþ1ðxÞ

¼ 2 (the only negative value). Thus,

TnðxÞ ffiffiffin p 1

happens innitely often as well. r We now complete the proof of the main theo- rem.

Proof of theorem 1. We rst note thatFaðxÞ has the following representation:

FaðxÞ ¼X1

k¼1

kðxÞ;

kðxÞ ¼Yk1

l¼1

p lðxÞ

q kðxÞ

; where

pð0Þ ¼a; pð1Þ ¼12a; pð2Þ ¼a;

qð0Þ ¼0; qð1Þ ¼a; qð2Þ ¼1a:

In what follows, we assume that a¼a0 and x satises

lim sup

n!1

TnðxÞ ffiffiffin p 1:

From the denition ofTnðxÞ, we can take an increas- ing sequencefrngwhich satises

Xrn

k¼1

c kðxÞ

pffiffiffiffiffirn

; rnþ1ðxÞ ¼1; n¼1;2;3; : Here we denefxngby

xn¼Xrn

k¼1

kðxÞ 3k : Then,

xxn¼ X1

k¼rnþ1

kðxÞ 3k 1

3rnþ1 >0

and we have the following evaluation:

FaðxÞ FaðxnÞ xxn

¼

P1

k¼rnþ1kðxÞ

P1

k¼rnþ1kðxÞ=3k

P1

k¼rnþ1kðxÞ

1=3rn

¼3rnYrn

l¼1

p lðxÞ

q

rnþ1ðxÞ

þ X1

k¼rnþ2

Y

k1

l¼rnþ1

p lðxÞ

q kðxÞ

3rnYrn

l¼1

p lðxÞ

q

rnþ1ðxÞ

X1

k¼rnþ2

Y

k1

l¼rnþ1

p lðxÞ

q

kðxÞ!

3rnYrn

l¼1

p lðxÞ

qð1Þ X1

k¼1

pð1Þmax

0l2

pðlÞk1max

0l2qðlÞ

!

¼3rnYrn

l¼1

p lðxÞ

a ð2a1ÞX1

k¼1

ak

!

¼að23aÞ

1a exp Xrn

l¼1

log 3p lðxÞ

! : Using the following relations:

log

3pð0Þ¼log

3pð2Þ¼logð3aÞ;

log

3pð1Þ¼log

3ð12aÞ¼log27a254a3 9a2

¼log 1 9a2

¼ 2 logð3aÞ;

we obtain

FaðxÞ FaðxnÞ xxn

að23aÞ

1a exp logð3aÞXrn

l¼1

c

lðxÞ! að23aÞ

1a ð3aÞpffiffiffiffirn : It follows that

n!1lim

FaðxÞ FaðxnÞ xxn

¼ 1:

(5)

Namely, FaðxÞ is non-dierentiable at x. From the previous lemma, we know that

lim sup

n!1

TnðxÞ ffiffiffin p 1

holds almost everywhere in½0;1Þ, and so, we can con- clude that FaðxÞ is non-dierentiable almost every- where in½0;1Þ.

Acknowledgments. I would like to express my gratitude to Prof. H. Okamoto for informing me of his interesting study about continuous non- dierentiable functions and for encouraging me to write this article after I proved the case a¼a0. I would also like to thank an anonymous referee for letting me know a much shorter proof of Lemma 1 using the law of the iterated logarithm.

References

[ 1 ] N. Bourbaki,Functions of a real variable, Trans- lated from the 1976 French original by Philip Spain, Springer, Berlin, 2004.

[ 2 ] P. Hartman and A. Wintner, On the law of the iterated logarithm, Amer. J. Math. 63 (1941), no. 1, 169{176.

[ 3 ] H. Okamoto, A remark on continuous, nowhere dierentiable functions, Proc. Japan Acad. Ser.

A Math. Sci.81(2005), no. 3, 47{50.

[ 4 ] H. Okamoto and M. Wunsch, A geometric con- struction of continuous, strictly increasing sin- gular functions, Proc. Japan Acad. Ser. A Math. Sci.83(2007), no. 7, 114{118.

[ 5 ] F. W. Perkins, An Elementary Example of a Continuous Non-Dierentiable Function, Amer.

Math. Monthly34(1927), no. 9, 476{478.

参照

関連したドキュメント

We show that a discrete fixed point theorem of Eilenberg is equivalent to the restriction of the contraction principle to the class of non-Archimedean bounded metric spaces.. We

Beer introduced the problem of the global coincidence on C(X, Y ) for metric spaces, and proved that if the metric space Y contains a non trivial arc, than the above two

By correcting these mistakes, we find that parameters of the spherical function are rational with respect to parameters of the (generalized principal series) representation.. As

In this section we state our main theorems concerning the existence of a unique local solution to (SDP) and the continuous dependence on the initial data... τ is the initial time of

CHANDRA, On the degree of approximation of a class of functions by means of Fourier series, Acta Math. CHANDRA, A note on the degree of approximation of continuous functions,

The construction extends the case of a group: the space of continuous functions with compact support on the groupoid is made into a ∗ -algebra and endowed with the smallest C ∗

For a complete valuation field k and a topological space X, we prove the universality of the underlying topological space of the Berkovich spectrum of the Banach k-algebra C bd (X,

As in the case of real-valued AC functions, the standard arc-length L(γ) of absolutely continuous curves γ as defined in (2.1) is well-defined on manifolds with smooth (or