Next Article in Journal
Investigating the Impact of COVID-19 on E-Learning: Country Development and COVID-19 Response
Previous Article in Journal
Quantile Regression with a New Exponentiated Odd Log-Logistic Weibull Distribution
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

On the Fractal Measures and Dimensions of Image Measures on a Class of Moran Sets

by
Najmeddine Attia
1,2,*,† and
Bilel Selmi
2,†
1
Department of Mathematics and Statistics, College of Science, King Faisal University, P.O. Box 400, Al-Ahsa 31982, Saudi Arabia
2
Analysis, Probability and Fractals Laboratory LR18ES17, Department of Mathematics, Faculty of Sciences of Monastir, University of Monastir, Monastir 5000, Tunisia
*
Author to whom correspondence should be addressed.
These authors contributed equally to this work.
Mathematics 2023, 11(6), 1519; https://doi.org/10.3390/math11061519
Submission received: 13 February 2023 / Revised: 3 March 2023 / Accepted: 13 March 2023 / Published: 21 March 2023

Abstract

:
In this work, we focus on the centered Hausdorff measure, the packing measure, and the Hewitt–Stromberg measure that determines the modified lower box dimension Moran fractal sets. The equivalence of these measures for a class of Moran is shown by having a strong separation condition. We give a sufficient condition for the equality of the Hewitt–Stromberg dimension, Hausdorff dimension, and packing dimensions. As an application, we obtain some relevant conclusions about the Hewitt–Stromberg measures and dimensions of the image measure of a τ -invariant ergodic Borel probability measures. Moreover, we give some statistical interpretation to dimensions and corresponding geometrical measures.

1. Introduction

The measure theory has an important role in formulating calculus on fractal sets [1,2,3,4,5,6,7]. The most important (and well-known) measures in fractal geometry are the Hausdorff measure, the packing measure, and the Hewitt–Stromberg measure. These measures have been investigated by several authors, highlighting their importance in the study of local properties of fractals and products of fractals [8,9,10,11,12,13,14,15,16,17,18,19]. Such measures appear explicitly, for example, in Pesin’s monograph and implicitly in Mattila’s text [20,21]. Finer measures have been used in the built analysis on a wider class of fractal sets [1]. Fractal Cantor-like sets and their measures were investigated in [7]. The quasi-Lipschitz equivalence of the Moran fractals was studied utilizing the Hausdorff dimension [22]. One of the purposes of this paper is to define and study a class of the Hewitt-Stromberg measures on Moran fractal sets. While Hausdorff and packing measures are defined using coverings and packing by families of sets with diameters that are less than a given positive number δ , say, the Hewitt-Stromberg measures are defined using ’packing of balls’ with a fixed diameter δ .
A Mathematical Theory of Communication, Claude E. Shannon’s 1948 article [23], was the first to establish the idea of entropy. Entropy is “a measure of the uncertainty associated with a random variable”, according to Wikipedia. In this context, “message” refers to a particular realization of the random variable, and “term” usually refers to the Shannon entropy, which measures the anticipated value of the information contained in a message, typically in unit-like bits. The Shannon entropy, on the other hand, measures the average amount of information that is lost when one does not know the value of the random variable. The introduction of Shannon entropy can be regarded as one of the most significant achievements during the previous fifty years in the literature on probabilistic uncertainty. Entropy created the framework for the thorough knowledge of communication theory. Several academic fields, such as statistical thermodynamics, spectral analysis, urban and regional planning, image reconstruction, business, queuing theory, economics, finance, operations research, biology, and manufacturing, among others, have used the concept of entropy. These applications will be discussed in the following section. This section reviews entropy as well as the related ideas of maximum entropy and directed divergence.
The Kolmogorov entropy is a crucial metric for describing how chaotic a system is. With relation to the phase point’s location on the attractor, it provides the average rate of information loss. The Rényi entropy [24] can be used to calculate the Kolmogorov entropy. The Shannon entropy in information theory is a particular instance of Rényi entropy. The thermodynamic entropy is the result of the Shannon entropy and Boltzmann constant. The fractal dimension is what distinguishes fractal formations. The family of fractal dimensions is limitless. In an e-dimensional space, a generalized fractal dimension can be defined. There is a direct relationship between the Rényi entropy and the generalized fractal dimension. The main purpose of this paper is to study some formulas for some fractal dimensions and measures of the image of measures that are computed using entropy on some Moran sets.
In the present paper, we investigate a class of Moran sets in general metric spaces and we discuss some proprieties and the equivalence of the fractal measures on these sets. As an application of the main result, we obtain similar formulas of measures and dimensions of the image of τ -invariant measures in symbolic space using entropy as in the classical case of self-similar sets. We give some interesting examples; in particular, we discuss a group of Moran sets and provide some statistical explanations for the dimensions and associated geometrical measures.
The outline of the paper is as follows: In the next section, we define modified Hausdorff and packing measures. In Section 3, Moran fractal sets are defined and strong separation conditions are given. Moreover, the equivalence of the Hausdorff measure, the packing measure, and the Hewitt-Stromberg measures are proved. Section 4 gives some results about the Hewitt–Stromberg measures and dimensions of the images of τ -invariant ones in symbolic space using entropy. Section 5 presents a conclusion.

2. The Fractal Measures

We first recall the definition of the centered Hausdorff measure, the packing measure, and the Hewitt–Stromberg measure. Throughout this paper, B ( x , r ) stands for the closed ball
B ( x , r ) = y R n | d ( x , y ) r .
For t 0 , E R n and δ > 0 , we define the packing pre-measure,
P ¯ t ( E ) = inf δ > 0 sup i ( 2 r i ) t | B ( x i , r i ) i is a centered δ - packing of E .
In a similar way, we define the Hausdorff pre-measure,
H ¯ t ( E ) = sup δ > 0 inf i ( 2 r i ) t | B ( x i , r i ) i is a centered δ - covering of E .
Function H ¯ t is σ -sub-additive but not increasing, and function P ¯ t is increasing but not σ -sub-additive. That is why we introduce the modifications of the Hausdorff and packing measures H t and P t :
H t ( E ) = sup F E H ¯ t ( F ) and P t ( E ) = inf E i E i i P ¯ t ( E i ) | E i R n , i .
The Hewitt–Stromberg pre-measure of E is defined by D t ( E ) = lim inf r 0 N r ( E ) ( 2 r ) t , where N r ( E ) is the minimum number of closed balls with diameter r, needed to cover E (the largest number of disjoint balls of radius r with centers in E). The Hewitt–Stromberg measure of E is defined by
d t ( E ) = inf i = 1 + D t ( E i ) | E i E i and the E i s are bounded .
The functions H t , P t , and d t are metric outer measures and, thus, measures on the Borel family of subsets of R n (see [15,16,25,26,27,28,29]). An important feature of the Hausdorff measure, packing measure, and Hewitt–Stromberg measure is that
H t ( E ) d t ( E ) P t ( E ) for any bounded set E R n .
The Hausdorff dimension, the packing dimension, and the Hewitt–Stromberg dimension (modified lower box dimension) can be defined by
dim H ( E ) = inf { t 0 H t ( E ) = 0 = sup { t 0 H t ( E ) = + , dim P ( E ) = inf { t 0 P t ( E ) = 0 = sup { t 0 P t ( E ) = + , dim ̲ M B ( E ) = inf { t 0 d t ( E ) = 0 = sup { t 0 d t ( E ) = + .
It follows immediately from the definitions that
dim H ( E ) dim ̲ M B ( E ) dim ¯ M B ( E ) = dim P ( E ) ,
where dim ¯ M B is the modified upper box dimension (see for example [1,15,29,30]).

3. The Equivalence of Fractal Measures on Moran Fractal Sets

We will start by defining the Moran sets. Let { n k } k and { Φ k } k 1 be, respectively, two sequences of positive integers and positive vectors, such that
Φ k = ( c k 1 , c k 2 , , c k n k ) , j = 1 n k c k j 1 , k N .
For any m , k N , such that m k , let
D m , k = { ( i m , i m + 1 , , i k ) 1 i j n j , m j k
and
D k = D 1 , k = { ( i 1 , i 2 , , i k ) 1 i j n j , 1 j k .
We also set D 0 = and D = k 0 D k , Considering σ = ( i 1 , i 2 , , i k ) D k , τ = ( j 1 , j 2 , , j m ) D k + 1 , m , we set
σ τ = ( i 1 , i 2 , , i k , j 1 , j 2 , , j m ) .
Definition 1 
([31,32]). Let X be a complete metric space and I X a compact set with no empty interior (for convenience, we assume that the diameter of I is 1). The collection F = { I σ | σ D } of subsets of I is called having Moran structure if
1.
For any ( i 1 , i 2 , , i k ) D k , I i 1 i 2 i k is similar to I. That is, there exists a similar transformation
S i 1 i 2 i k : X X I I i 1 i 2 i k ,
where we assume that I = I .
2.
For all k 1 , ( i 1 , i 2 , , i k 1 ) D k 1 , I i 1 i 2 i k i k { 1 , 2 , , n k } are subsets of I i 1 i 2 i k 1 and
I i 1 i 2 i k 1 , i k I i 1 i 2 i k 1 , i k = , 1 i k < i k n k ,
where I denotes the interior of I.
3.
For all k 1 and 1 j n k , taking ( i 1 , i 2 , , i k 1 , j ) D k , we have
0 < c k j = c i 1 i 2 i k 1 j = | I i 1 i 2 i k 1 j | | I i 1 i 2 i k 1 | < 1 , k 2 ,
where | I | denotes the diameter of I.
Suppose that F is a collection of subsets of I having Moran structure. We call E = k 1 σ D k I σ , a Moran set determined by F , and call F k = I σ , σ D k the k-order fundamental sets of E. I is called the original set of E. We assume lim k + max σ D k | I σ | = 0 . Then, for all i D , the set n 1 I i 1 i 2 i n is a single point. We shall denote it by φ ( i ) . For all w = ( i 1 , i 2 , i k , ) D , we use the abbreviation w | k for the first k elements of the sequence,
I k ( w ) = I w | k = I i 1 i 2 i k , and c n ( w ) = c i 1 i 2 i n .
Here, we consider a class of Moran sets E which satisfy a special property called the strong separation condition (SSC), i.e., Take any I σ F . Let I σ * 1 , I σ * 2 , , I σ * n k + 1 be the ( k + 1 ) -order fundamental subsets. We say that I σ satisfies the (SSC) if dist ( I σ * i , I σ * j ) δ k | I σ | , for all i j , where ( δ k ) k is a sequence of positive real numbers, such that 0 < δ = inf k δ k < 1 .
We focus on the equivalence of the centered Hausdorff measure, the packing measure, and the Hewitt–Stromberg measure for the Moran sets meeting the strong separation requirement. We outline the resources and preliminary findings that will be used in the validation of our major findings.
Lemma 1. 
Let μ be a finite Borel measure on R n and E be a bounded Borel set. Then there exist two positive constants A 1 , A 2 , such that
A 1 d t ( E ) inf x E lim inf r 0 μ B ( x , r ) r t μ ( E ) A 2 d t ( E ) sup x E lim sup r 0 μ B ( x , r ) r t .
Proof. 
Follows directly from Theorem 2.1 in [32].    □
Lemma 2 
([31,32]). Suppose that E I is a Moran set satisfying (SSC) and let μ be a finite Borel measure, such that supp μ E . Then, there exist positive constants c i , 1 i 4 relating to δ , t , with which the following inequalities hold for any φ ( i ) E
c 1 lim inf n + μ ( I n ( i ) ) | I n ( i ) | t lim inf r 0 μ B ( φ ( i ) , r ) r t c 2 lim inf n + μ ( I n ( i ) ) | I n ( i ) | t
and
c 3 lim sup n + μ ( I n ( i ) ) | I n ( i ) | t lim sup r 0 μ B ( φ ( i ) , r ) r t c 4 lim sup n + μ ( I n ( i ) ) | I n ( i ) | t .
Let μ be a finite Borel measure on R n , we define the dimension of a measure μ by
dim ̲ M B ( μ ) = sup { t 0 d t ( μ ) = + = inf { t 0 d t ( μ ) = 0 ,
where
d t ( μ ) = inf E R n { d t ( E ) μ ( R n E ) = 0 .
Definition 2. 
We say that two Borel measures μ and ν are equivalent and we write μ ν if for any Borel set A, we have μ ( A ) = 0 ν ( A ) = 0 .
Theorem 1. 
Suppose that E I is a Moran set satisfying (SSC). Let μ be a finite Borel measure such that supp μ E .
1.
Suppose that there exists γ, such that
lim n + μ ( I n ( i ) ) | I n ( i ) | t = 0 i f t < γ , + i f t > γ , for any i D .
Then, dim ̲ M B ( E ) = γ = dim ̲ M B ( μ ) .
2.
Suppose that μ satisfies a stronger condition at γ, i.e.,
0 < lim n + μ ( I n ( i ) ) | I n ( i ) | γ < + , for any i D .
Then, μ d γ on E .
Proof. 
1.
From Lemmas 1 and 2, we can see that if t > γ , then
d t ( E ) A 1 1 μ ( E ) sup x E lim sup r 0 r t μ B ( x , r ) C μ ( E ) sup i D lim sup n + | I n ( i ) | t μ I n ( i ) = C μ ( E ) sup i D lim inf n + μ I n ( i ) | I n ( i ) | t 1 = 0 ,
where C = A 1 1 c 2 1 . It follows that dim ̲ M B ( E ) γ .
On the other hand, we can also see that if t < γ , then for all set F E such that μ ( F ) > 0 , we have
d t ( F ) A 2 1 μ ( F ) inf x F lim inf r 0 r t μ B ( x , r ) C 1 μ ( F ) inf i φ 1 ( F ) lim inf n + | I n ( i ) | t μ I n ( i ) = C 1 μ ( F ) inf i φ 1 ( F ) lim sup n + μ I n ( i ) | I n ( i ) | t 1 = + ,
where C 1 = A 2 1 c 4 1 . This leads to dim ̲ M B ( μ ) γ . Thus, dim ̲ M B ( μ ) dim ̲ M B ( E ) . From dim ̲ M B ( μ ) dim ̲ M B ( E ) , we have dim ̲ M B ( μ ) = γ = dim ̲ M B ( E ) .
2.
Suppose that μ satisfies (2) and set
Ω n = i D | lim sup r 0 r γ μ B ( φ ( i ) , r ) < n .
Then D n 1 Ω n . Suppose that μ ( B ) = 0 , for any B E . Then, there exists a sequence of open sets { G k } k , such that B G k and μ ( G k ) 1 k , for all k . Putting μ k ( A ) = μ ( A G k ) and taking into account that
lim sup r 0 r γ μ k B ( x , r ) = lim sup r 0 r γ μ B ( x , r ) , for all x B .
From Lemma 1, we can see that, for any k , n N and x B , we have
d t B φ ( Ω n ) A 1 1 μ k ( B ) sup x B φ ( Ω n ) lim sup r 0 r γ μ k B ( x , r ) A 1 1 μ ( G k ) sup x B φ ( Ω n ) lim sup r 0 r γ μ B ( x , r ) A 1 1 n k .
Letting k + , we obtain d t B φ ( Ω n ) = 0 , for any n N . This leads to
d t ( B ) n 1 d t B φ ( Ω n ) = 0 .
On the other hand, if we assume that d t ( B ) = 0 , for all B E , then setting
Ω n = i D | lim inf r 0 r γ μ B ( φ ( i ) , r ) > 1 n ,
we easily obtain, from Lemma 1, where
0 = d t B φ ( Ω n ) A 2 1 μ B φ ( Ω n ) inf x B φ ( Ω n ) lim inf r 0 r γ μ B ( x , r ) A 2 1 n μ B φ ( Ω n ) .
This implies that μ B φ ( Ω n ) = 0 , for any n N . Now, thanks to the fact that μ φ n 1 Ω n = 1 , we deduce that μ ( B ) = 0 , which ends the proof.
   □
Example 1. 
We will consider in this example a special case in which the conditions of all the numbers c i 1 , i 2 , , i n depend only on the length and the last variable, i.e.,
c i 1 , i 2 , , i n = c n , i n .
We define the pressure function π and the Gibbs measure μ t defined respectively by
π ( t ) = lim n + π n ( t ) where π n ( t ) = 1 n log ( i 1 , i 2 , , i n ) D n j = 1 n c j , i j t .
and
μ t ( i 1 i 2 i n ) = j = 1 n c j , i j t Z j t ,
where Z n t = j c n , j t for any ( i 1 , i 2 , , i n ) D n , n N . It is not difficult to see that π is strictly decreasing and continuous. Moreover, we have
μ t ( i 1 i 2 i n ) = | I n ( w ) | t j = 1 n Z j t 1 .
Since, for all n, we have
( i 1 , i 2 , i n ) D n j = 1 n c j , i j t = j = 1 n Z j t
then
lim n + μ t ( i 1 i 2 i n ) | I n ( w ) | γ = lim n + | I n ( w ) | t γ e log j = 1 n Z j t = lim n + | I n ( w ) | t γ e n π n ( t ) .
Now, suppose that t : = α is the unique number, such that
( i 1 , i 2 , , i n ) D n | I i 1 i 2 i n | α = 1 .
It is clear that π ( α ) = 0 , and then
lim n + μ α ( i 1 i 2 i n ) | I n ( w ) | γ = + if γ > α
and
lim n + μ α ( i 1 i 2 i n ) | I n ( w ) | γ = 0 if γ < α .
It follows from Theorem 1 that α = dim ̲ M B ( E ) = dim ̲ M B ( μ α ) .
Theorem 2. 
Suppose that E I is a Moran set satisfying (SSC). Let μ be a finite Borel measure with supp μ E and 0 < α < n such that
0 < lim n + μ I n ( i ) | I n ( i ) | α < + for any i D .
Then
dim H ( E ) = dim ̲ M B ( E ) = dim ¯ M B ( E ) = dim P ( E ) = α .
In addition,
H α d α P α on E .
Proof. 
Follows from Theorem 1 and ([32], Theorems 4.2 and 4.3). □
Example 2. 
Let I = [ 0 , 1 ] , n k = 2 and c k j = 1 3 far all k 1 and 1 j n k . In this case, the Moran set E is the classical ternary Cantor set. Let α = log 2 log 3 and μ be a probability measure on I such that
μ I n ( i ) = | I n ( i ) | α if i D , 0 otherwise .
It is clear that supp μ E and
lim n + μ I n ( i ) | I n ( i ) | α = 1 .
Therefore,
dim H ( E ) = dim ̲ M B ( E ) = dim ¯ M B ( E ) = dim P ( E ) = log 2 log 3 and H α d α P α on E .
Example 3. 
Let A = { a , b } be a two-letter alphabet, and A * the free monoid generated by A. Let F be the homomorphism on A * , defined by F ( a ) = a b and F ( b ) = a . It is easy to see that F n ( a ) = F n 1 ( a ) F n 2 ( a ) . We denote by F n ( a ) the length of the word F n ( a ) , thus
F n ( a ) = s 1 s 2 s F n ( a ) , s i A .
Therefore, as n + , we have the infinite sequence
ω = lim n + F n ( a ) = s 1 s 2 s 3 s n { a , b } N ,
which is called the Fibonacci sequence. For any n 1 , write ω n = ω n = s 1 s 2 s n . We denote by ω n a the number of the occurrence of the letter a in ω n , and ω n b the number of occurrences of b . Then, ω n a + ω n b = n . It follows that lim n + ω n a n = η , where η 2 + η = 1 .
Let 0 < r a < 1 2 , 0 < r b < 1 3 , r a , r b R . In the Moran construction above, let
n k = 2 , if s k = a 3 , if s k = b ,
c k j = c k = r a , if s k = a r b , if s k = b , 1 j n k .
Then we construct the homogeneous Moran set related to the Fibonacci sequence and denote it by E : = E ( ω ) = I , n k , c k . Through the construction of E , we have
I σ = r a ω k a r b ω k b , σ D k .
There exists a probability measure ν α supported by E such that for any k 1 and σ 0 D k ,
ν α I σ 0 = I σ 0 α σ D k I σ α , for all α R .
Let α = log 2 η log 3 log r a + η log r b , where η 2 + η = 1 . It is clear that there exists a positive constant c, such that
lim n + ν α ( I n ( w ) ) | I n ( w ) | γ = c lim n + | I n ( w ) | α γ for all γ R .
This implies that
lim n + ν α ( I n ( w ) ) | I n ( w ) | γ = + if γ > α
and
lim n + ν α ( I n ( w ) ) | I n ( w ) | γ = 0 if γ < α .
Finally, Theorem 1 gives that dim ̲ M B ( E ) = dim ̲ M B ( ν α ) = log 2 η log 3 log r a + η log r b , where η 2 + η = 1 .
Remark 1. 
As we analyze the Hewitt–Stromberg measures and dimensions, our findings are closely comparable to those in the cited work [32], for which M. Dai also demonstrates similar findings for the Hausdorff and packing measures. Even though the results are parallel, we want to point out that the goal of our paper is to draw some useful inferences about the Hewitt–Stromberg measures and dimensions of the image measure of τ-invariant ergodic Borel probability measures, and provide a statistical explanation for the dimensions and associated geometrical measures in the next section.

4. The Dimension of Image Measure

In this section, we use entropy to derive analogous formulas for the Hewitt–Stromberg measures and the image dimensions of τ -invariant measures in a symbolic space. We focus on a particular class of Moran sets and provide some statistical explanations for the dimensions and associated geometrical measures.
Let E be a Moran set satisfying the strong separation condition and μ be a probability measure defined on the σ -algebra F of subsets of D. Wa said that τ preserves the measure μ (or μ is τ -invariant), if
μ ( τ 1 ( B ) ) = μ ( B ) , B F .
An invariant probability measure is said to be ergodic if for every set B F , τ 1 ( B ) = B has either zero or full measure.
Theorem 3 
([33]). Let μ be the τ-invariant ergodic probability measure on D and f L 1 ( D , R ) . Then
lim n + 1 n k = 0 n 1 f τ k = D f d μ , μ a . e .
We define the entropy h μ ( τ ) of τ with respect to the measure μ by
h μ ( τ ) = lim n + 1 n w D n μ ( i 1 i 2 , i n ) log μ ( i 1 i 2 i n ) .
In this section, we will study the image measure of μ by φ denoted by φ * μ and defined by
φ * μ ( B ) = μ φ 1 ( B ) , for all B B ( X )
where B ( X ) is the Borel algebra on X and φ : D E is a measurable map. We will assume through this section that
0 < lim inf n + j = 1 n c j ( w ) c j ( τ ( w ) ) lim sup n + j = 1 n c j ( w ) c j ( τ ( w ) ) < +
and
lim inf n + 1 n log | I n ( w ) | = lim sup n + 1 n log | I n ( w ) |
for all w = ( i 1 , i 2 , i k , ) D . Our main result in this section is the following.
Theorem 4. 
Let E be a Moran set satisfying the strong separation condition and μ be a τ-invariant ergodic Borel probability measures on D.
1.
One has
dim ̲ M B φ * μ = h μ ( τ ) lim n + 1 n j = 1 n ln c j ( i ) d μ 1 .
2.
If we denote by α = dim ̲ M B ( φ * μ ) then
d α ( φ * μ ) = 0 lim n + φ * μ ( I n ( w ) ) j = 1 n c j ( w ) α = + μ a . e . w .
0 < d α ( φ * μ ) < + 0 < lim n + φ * μ ( I n ( w ) ) j = 1 n c j ( w ) α < + μ a . e . w .
d α ( φ * μ ) = + lim n + φ * μ ( I n ( w ) ) j = 1 n c j ( w ) α = 0 μ a . e . w .
Proof. 
1.
We recall that | I n ( w ) | = j = 1 n c j ( w ) , for all w = ( i 1 , i 2 , i k , ) D and then using (6), we set
Γ + ( w ) = lim n + 1 n j = 1 n log c j ( w ) .
We consider, for α R , the set
D μ ( α ) = w D , Γ + ( w ) = α and lim n + 1 n log μ ( w | n ) = h μ ( τ ) .
From (5) we have Γ + ( w ) = Γ + ( τ ( w ) ) > 0 and since μ is ergodic, it follows from Theorem 3 that there exists α μ such that
Γ + ( w ) = α μ = Γ + ( w ) d μ , μ a . e . on D ,
and
lim n + 1 n log μ ( w | n ) = h μ ( τ ) μ a . e . ,
by Shannon–McMillan–Breiman Theorem. Therefore, μ ( D μ ( α μ ) ) = 1 , which implies by using (7), for ϵ > 0 that
j = 1 n c j ( w ) h μ ( τ ) / α μ e n h μ ( τ ) + n ϵ h μ ( τ ) / α μ , μ ( w | n ) e n h μ ( τ ) + n ϵ ,
for almost every w D μ ( α μ ) and n big enough. Now, we put B B ( X ) , such that φ 1 ( B ) = D μ ( α μ ) . For all w φ 1 ( B ) with B E , we have
φ * μ ( I n ( w ) ) j = 1 n c j ( w ) t e n ϵ ( 1 t ) 1 ,
for t = h μ ( τ ) α μ . By using Lemma 1, where there exists a positive constant A 1 , such that
A 1 d t ( B ) inf w φ 1 ( B ) lim inf n + φ * μ ( I n ( w ) ) j = 1 n c j ( w ) t φ * μ ( B )
which implies that
dim ̲ M B ( B ) h μ ( τ ) α μ and then dim ̲ M B ( φ * μ ) h μ ( τ ) α μ .
On the other hand, let B, such that φ * μ ( B ) = 1 and take C = φ 1 ( B ) D μ ( α μ ) . Since μ ( C ) = 1 , then, using again Lemma 1, we can deduce that
dim ̲ M B ( B ) dim ̲ M B ( C ) h μ ( τ ) α μ
and then we have equality using (8).
2.
For all w D , we set
B ( w ) = lim n + φ * μ ( I n ( w ) ) j = 1 n c j ( w ) α .
If we assume that d α ( φ * μ ) = 0 then, for each n N , there exists B n E , such that
φ * μ ( E B n ) = 0 and d α ( B n ) < 1 n .
Now consider the following sets
B 0 = w D , B ( w ) = 0 ,
B + = w D , 1 / k < B ( w ) < k
and
B = w D , B ( w ) = + .
Since μ is ergodic, these sets are either null or full with respect to the measure μ . Now, Suppose that μ ( B ) = 1 , where B { B 0 , B + , B } , and put Z n , k = B n φ ( B k ) where
B k = w D , 1 k B ( w ) < + if B = B + , B k = w D , B ( w ) = 0 if B = B 0 , B k = w D , B ( w ) = + if B = B .
It is clear that B k B as k + , which implies that lim k + φ * μ ( Z n , k ) = 1 , for any n N . Moreover,
d α ( Z n , k ) d α ( B n ) 1 / n , k N .
which gives, from Lemma 1, where there exists a positive constant A 2 , such that
φ * μ ( Z n , k ) A 2 d α ( Z n , k ) inf w φ 1 ( Z n , k ) B ( w ) .
Choose k large enough to satisfy φ * μ ( Z n , k ) > 1 / 2 . Then
1 / 2 A 2 d α ( Z n , k ) inf w φ 1 ( Z n , k ) B ( w ) A 2 n inf w φ 1 ( Z n , k ) B ( w ) ,
which creates a contradiction if B = B 0 . Now, if B = B + , then the above inequalities imply that
n 2 A 2 k ,
for any n and any sufficiently large k, which is not possible. Finally, we conclude that B = B .
The two other assertions may be proved in the same way. In what follows, we will only prove the second one and keep the other to the readers. We suppose conversely that B = B μ a.e. w. It follows from Lemma 1 that d α ( B ) = 0 whenever φ * μ ( B ) = 1 and then
d α ( φ * μ ) ) = 0 .
   □
Example 4. 
Consider again the special case studied in Example 1 in which the conditions of all the numbers c i 1 , i 2 , , i n depend only on the length and the last variable, i.e.,
c i 1 , i 2 , , i n = c n , i n .
Let α be the unique number, such that
( i 1 , i 2 , , i n ) D n | I i 1 i 2 i n | α = 1 ,
and μ α the associated Gibbs measure (see (4)). Then
φ * μ α ( I n ( w ) ) = μ α ( i 1 i 2 i n ) = | I n ( w ) | α j = 1 n Z j α 1 .
Recall the definition of the pressure function π (see (3)), then
lim n + φ * μ α ( I n ( w ) ) | I n ( w ) | γ = lim n + | I n ( w ) | α γ e log j = 1 n Z j α = lim n + | I n ( w ) | α γ e n π n ( α ) .
which implies that
lim n + φ * μ ( I n ( w ) ) | I n ( w ) | γ = + if γ > α
and
lim n + φ * μ ( I n ( w ) ) | I n ( w ) | γ = 0 if γ < α .
It follows from Theorem 1 that
α = dim ̲ M B ( E ) = dim ̲ M B ( φ * μ α ) .
Moreover, if we assume that
0 < lim n + ( i 1 , i 2 , , i n ) D n j = 1 n c j , i j α < +
then by using Theorems 1 and 4 we have
d α φ * μ α on E .

5. Conclusions

In this study, we focus on the characteristics of a class of Moran sets for the centered Hausdorff measure, the Packing measure, and the Hewitt–Stromberg measure. We primarily explore the equivalence of these measures for a class of Moran sets meeting the strong separation criterion. We specifically address a class of Moran sets in general metric spaces with respect to the equality of the three aforementioned dimensions and their accompanying measures. Using entropy, we derive measurements and dimensions of the image of τ -invariant measures in symbolic spaces that are analogous to those obtained in the traditional situation of self-similar sets. Moreover, we provide several intriguing examples, focusing in particular on a class of Moran sets and providing some statistical explanations for dimensions and corresponding geometrical measures.

Author Contributions

Writing—review & editing, N.A. and B.S. Investigation, N.A and B.S. All authors have read and agreed to the published version of the manuscript.

Funding

The authors extend their appreciation to the Deanship of Scientific Research, Vice Presidency for Graduate Studies and Scientific Research at King Faisal University, Saudi Arabia, for financial support under the annual funding track [GRANT3069].

Data Availability Statement

All data generated or analyzed during this study are included in this published article.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Falconer, K.J. Fractal Geometry: Mathematical Foundations and Applications; Wiley: Hoboken, NJ, USA, 1990. [Google Scholar]
  2. WithersReichert, W.D. Fundamental theorems of calculus for Hausdorff measures on the real line. J. Math. Anal. Appl. 1988, 129, 581–595. [Google Scholar]
  3. Jiang, H.; Su, W. Some fundamental results of calculus on fractal sets. Commun. Nonlinear Sci. Numer. Simul. 1998, 3, 22–26. [Google Scholar] [CrossRef]
  4. Wu, J.; Wang, C. Fractal Stokes’ Theorem Based on Integration on Fractal Manifolds. Fractals 2020, 28, 2050010. [Google Scholar] [CrossRef]
  5. Kesseböhmer, M.; Samuel, T.; Weyer, H. A note on measure-geometric Laplacians. Mon. Math. 2016, 181, 643–655. [Google Scholar] [CrossRef] [Green Version]
  6. Kunze, H.; La Torre, D.; Mendivil, F.; Vrscay, E.R. Self-similarity of solutions to integral and differential equations with respect to a fractal measure. Fractals 2019, 27, 1950014. [Google Scholar] [CrossRef]
  7. Golmankhaneh, A.K.; Fernez, A.; Golmankhaneh, A.K.; Baleanu, D. Diffusion on middle-ε Cantor sets. Entropy 2018, 20, 504. [Google Scholar] [CrossRef] [Green Version]
  8. Hewitt, E.; Stromberg, K. Real and Abstract Analysis. A Modern Treatment of the Theory of Functions of a Real Variable; Springer: Berlin/Heidelberg, Germany, 1965. [Google Scholar]
  9. An, L.X.; He, X.G. A class of spectral Moran measure. J. Funct. Anal. 2014, 266, 343–354. [Google Scholar] [CrossRef]
  10. Haase, H. A contribution to measure and dimension of metric spaces. Math. Nachr. 1985, 124, 45–55. [Google Scholar] [CrossRef]
  11. Haase, H. Open-invariant measures and the covering number of sets. Math. Nachr. 1987, 134, 295–307. [Google Scholar] [CrossRef]
  12. Haase, H. The dimension of analytic sets. Acta Universitatis Carolinae. Math. Phys. 1988, 29, 15–18. [Google Scholar]
  13. Haase, H. Dimension functions. Math. Nachr. 1989, 141, 101–107. [Google Scholar] [CrossRef]
  14. Haase, H. Fundamental theorems of calculus for packing measures on the real line. Math. Nachr. 1990, 148, 293–302. [Google Scholar] [CrossRef]
  15. Jurina, S.; MacGregor, N.; Olsen, L.; Stylianou, A. On the Hausdorff and packing measures of typical compact metric spaces. Aequ. Math. 2018, 92, 709–735. [Google Scholar] [CrossRef] [Green Version]
  16. Olsen, L. On average Hewitt--Stromberg measures of typical compact metric spaces. Math. Z. 2019, 293, 1201–1225. [Google Scholar] [CrossRef] [Green Version]
  17. Zindulka, O. Packing measures and dimensions on Cartesian products. Publ. Mat. 2013, 57, 393–420. [Google Scholar] [CrossRef] [Green Version]
  18. Attia, N.; Selmi, B. Regularities of multifractal Hewitt--Stromberg measures. Commun. Korean Math. Soc. 2019, 34, 213–230. [Google Scholar]
  19. Edgar, G.A. Integral, Probability, and Fractal Measures; Springer: Berlin/Heidelberg, Germany, 1998. [Google Scholar]
  20. Pesin, Y. Dimension Theory in Dynamical Systems, Contemporary Views and Applications; University of Chicago Press: Chicago, IL, USA, 1997. [Google Scholar]
  21. Mattila, P. Geometry of Sets and Measures in Euclidian Spaces: Fractals and Rectifiability; Cambridge University Press: Cambridge, UK, 1995. [Google Scholar]
  22. Li, H.; Wang, L.; Xi, L. Classification of Moran fractals. J. Math. Anal. Appl. 2011, 378, 230–237. [Google Scholar] [CrossRef] [Green Version]
  23. Shannon, C.E. A Mathematical Theory of Communication. Bell Syst. Tech. J. 1948, 27, 379–423. [Google Scholar] [CrossRef] [Green Version]
  24. Rényi, A. On Measures of Entropy and Information. Berkeley Symp. on Math. Statist. Prob. 1961, 1, 547–561. [Google Scholar]
  25. Attia, N.; Guedri, G. A note on the Regularities of Hewitt–Stromberg h-measures. Ann. Univ. Ferrara 2022. to appear. [Google Scholar] [CrossRef]
  26. Guedri, R.; Attia, N. A note on the generalized Hausdorff and packing measures of product sets in metric space. Math. Inequal. Appl. 2022, 25, 335–358. [Google Scholar] [CrossRef]
  27. Attia, N.; Jebali, H.; Guedri, G. On a class of Hausdorff measure of cartesian sets in metric spaces. Topol. Methods Nonlinear Anal. 2023. to appear. [Google Scholar]
  28. Baek, H.K.; Lee, H.H. Regularities of Fractal Measures on R2. Acta Math. Hung. 2003, 99, 25–32. [Google Scholar] [CrossRef]
  29. Lee, H.H.; Baek, I.S. The comparison of d-meuasure with packing and Hausdorff measures. Kyungpook Math. J. 1992, 32, 523–531. [Google Scholar]
  30. Mahjoub, A.; Attia, N. A relative vectorial multifractal formalism. Chaos Solitons Fractals 2022, 160, 112221. [Google Scholar] [CrossRef]
  31. Attia, N.; Selmi, B.; Souissi, C. Some density results of relative multifractal analysis. Chaos Solitons Fract. 2017, 103, 1–11. [Google Scholar] [CrossRef]
  32. Dai, M. The equivalence of measures on Moran set in general metric space. Chaos Solitons Fract. 2006, 29, 55–64. [Google Scholar] [CrossRef]
  33. Bedford, T.; Keane, M.; Series, C. Ergodic Theory, Symbolic Dynamics, and Hyperbolic Spaces; Oxford University Press: New York, NY, USA, 1991. [Google Scholar]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Attia, N.; Selmi, B. On the Fractal Measures and Dimensions of Image Measures on a Class of Moran Sets. Mathematics 2023, 11, 1519. https://doi.org/10.3390/math11061519

AMA Style

Attia N, Selmi B. On the Fractal Measures and Dimensions of Image Measures on a Class of Moran Sets. Mathematics. 2023; 11(6):1519. https://doi.org/10.3390/math11061519

Chicago/Turabian Style

Attia, Najmeddine, and Bilel Selmi. 2023. "On the Fractal Measures and Dimensions of Image Measures on a Class of Moran Sets" Mathematics 11, no. 6: 1519. https://doi.org/10.3390/math11061519

APA Style

Attia, N., & Selmi, B. (2023). On the Fractal Measures and Dimensions of Image Measures on a Class of Moran Sets. Mathematics, 11(6), 1519. https://doi.org/10.3390/math11061519

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop