# 分形维数

Example of non-integer dimensions. The first four iterations of the Koch curve, where after each iteration, all original line segments are replaced with four, each a self-similar copy that is 1/3 the length of the original. One formalism of the Hausdorff dimension uses this scale factor (3) and the number of self-similar objects (4) to calculate the dimension, D, after the first iteration to be D = (log N)/(log S) = (log 4)/(log 3) ≈ 1.26.[1] That is, while the Hausdorff dimension of a single point is zero, of a line segment is 1, of a square is 2, and of a cube is 3, for fractals such as this, the object can have a non-integer dimension.

Example of non-integer dimensions. The first four iterations of the Koch curve, where after each iteration, all original line segments are replaced with four, each a self-similar copy that is 1/3 the length of the original. One formalism of the Hausdorff dimension uses this scale factor (3) and the number of self-similar objects (4) to calculate the dimension, D, after the first iteration to be D = (log N)/(log S) = (log 4)/(log 3) ≈ 1.26. That is, while the Hausdorff dimension of a single point is zero, of a line segment is 1, of a square is 2, and of a cube is 3, for fractals such as this, the object can have a non-integer dimension.

In mathematics, Hausdorff dimension is a measure of roughness, or more specifically, fractal dimension, that was first introduced in 1918 by mathematician Felix Hausdorff.[2] For instance, the Hausdorff dimension of a single point is zero, of a line segment is 1, of a square is 2, and of a cube is 3. That is, for sets of points that define a smooth shape or a shape that has a small number of corners—the shapes of traditional geometry and science—the Hausdorff dimension is an integer agreeing with the usual sense of dimension, also known as the topological dimension. However, formulas have also been developed that allow calculation of the dimension of other less simple objects, where, solely on the basis of their properties of scaling and self-similarity, one is led to the conclusion that particular objects—including fractals—have non-integer Hausdorff dimensions. Because of the significant technical advances made by Abram Samoilovitch Besicovitch allowing computation of dimensions for highly irregular or "rough" sets, this dimension is also commonly referred to as the Hausdorff–Besicovitch dimension.

In mathematics, Hausdorff dimension is a measure of roughness, or more specifically, fractal dimension, that was first introduced in 1918 by mathematician Felix Hausdorff. For instance, the Hausdorff dimension of a single point is zero, of a line segment is 1, of a square is 2, and of a cube is 3. That is, for sets of points that define a smooth shape or a shape that has a small number of corners—the shapes of traditional geometry and science—the Hausdorff dimension is an integer agreeing with the usual sense of dimension, also known as the topological dimension. However, formulas have also been developed that allow calculation of the dimension of other less simple objects, where, solely on the basis of their properties of scaling and self-similarity, one is led to the conclusion that particular objects—including fractals—have non-integer Hausdorff dimensions. Because of the significant technical advances made by Abram Samoilovitch Besicovitch allowing computation of dimensions for highly irregular or "rough" sets, this dimension is also commonly referred to as the Hausdorff–Besicovitch dimension.

The Hausdorff dimension, more specifically, is a further dimensional number associated with a given set, where the distances between all members of that set are defined. Such a set is termed a metric space. The dimension is drawn from the extended real numbers, $\displaystyle{ \overline{\mathbb{R}} }$, as opposed to the more intuitive notion of dimension, which is not associated to general metric spaces, and only takes values in the non-negative integers.

The Hausdorff dimension, more specifically, is a further dimensional number associated with a given set, where the distances between all members of that set are defined. Such a set is termed a metric space. The dimension is drawn from the extended real numbers, $\displaystyle{ \overline{\mathbb{R}} }$, as opposed to the more intuitive notion of dimension, which is not associated to general metric spaces, and only takes values in the non-negative integers.

In mathematical terms, the Hausdorff dimension generalizes the notion of the dimension of a real vector space. That is, the Hausdorff dimension of an n-dimensional inner product space equals n. This underlies the earlier statement that the Hausdorff dimension of a point is zero, of a line is one, etc., and that irregular sets can have noninteger Hausdorff dimensions. For instance, the Koch snowflake shown at right is constructed from an equilateral triangle; in each iteration, its component line segments are divided into 3 segments of unit length, the newly created middle segment is used as the base of a new equilateral triangle that points outward, and this base segment is then deleted to leave a final object from the iteration of unit length of 4.[3] That is, after the first iteration, each original line segment has been replaced with N=4, where each self-similar copy is 1/S = 1/3 as long as the original.[1] Stated another way, we have taken an object with Euclidean dimension, D, and reduced its linear scale by 1/3 in each direction, so that its length increases to N=SD.[4] This equation is easily solved for D, yielding the ratio of logarithms (or natural logarithms) appearing in the figures, and giving—in the Koch and other fractal cases—non-integer dimensions for these objects.

In mathematical terms, the Hausdorff dimension generalizes the notion of the dimension of a real vector space. That is, the Hausdorff dimension of an n-dimensional inner product space equals n. This underlies the earlier statement that the Hausdorff dimension of a point is zero, of a line is one, etc., and that irregular sets can have noninteger Hausdorff dimensions. For instance, the Koch snowflake shown at right is constructed from an equilateral triangle; in each iteration, its component line segments are divided into 3 segments of unit length, the newly created middle segment is used as the base of a new equilateral triangle that points outward, and this base segment is then deleted to leave a final object from the iteration of unit length of 4. That is, after the first iteration, each original line segment has been replaced with N=4, where each self-similar copy is 1/S = 1/3 as long as the original. This equation is easily solved for D, yielding the ratio of logarithms (or natural logarithms) appearing in the figures, and giving—in the Koch and other fractal cases—non-integer dimensions for these objects.

The Hausdorff dimension is a successor to the simpler, but usually equivalent, box-counting or Minkowski–Bouligand dimension.

The Hausdorff dimension is a successor to the simpler, but usually equivalent, box-counting or Minkowski–Bouligand dimension.

## Intuition概念

The intuitive concept of dimension of a geometric object X is the number of independent parameters one needs to pick out a unique point inside. However, any point specified by two parameters can be instead specified by one, because the cardinality of the real plane is equal to the cardinality of the real line (this can be seen by an argument involving interweaving the digits of two numbers to yield a single number encoding the same information). The example of a space-filling curve shows that one can even map the real line to the real plane surjectively (taking one real number into a pair of real numbers in a way so that all pairs of numbers are covered) and continuously, so that a one-dimensional object completely fills up a higher-dimensional object.

The intuitive concept of dimension of a geometric object X is the number of independent parameters one needs to pick out a unique point inside. However, any point specified by two parameters can be instead specified by one, because the cardinality of the real plane is equal to the cardinality of the real line (this can be seen by an argument involving interweaving the digits of two numbers to yield a single number encoding the same information). The example of a space-filling curve shows that one can even map the real line to the real plane surjectively (taking one real number into a pair of real numbers in a way so that all pairs of numbers are covered) and continuously, so that a one-dimensional object completely fills up a higher-dimensional object.

Every space filling curve hits some points multiple times, and does not have a continuous inverse. It is impossible to map two dimensions onto one in a way that is continuous and continuously invertible. The topological dimension, also called Lebesgue covering dimension, explains why. This dimension is n if, in every covering of X by small open balls, there is at least one point where n + 1 balls overlap. For example, when one covers a line with short open intervals, some points must be covered twice, giving dimension n = 1.

Every space filling curve hits some points multiple times, and does not have a continuous inverse. It is impossible to map two dimensions onto one in a way that is continuous and continuously invertible. The topological dimension, also called Lebesgue covering dimension, explains why. This dimension is n if, in every covering of X by small open balls, there is at least one point where n + 1 balls overlap. For example, when one covers a line with short open intervals, some points must be covered twice, giving dimension n = 1.

But topological dimension is a very crude measure of the local size of a space (size near a point). A curve that is almost space-filling can still have topological dimension one, even if it fills up most of the area of a region. A fractal has an integer topological dimension, but in terms of the amount of space it takes up, it behaves like a higher-dimensional space.

But topological dimension is a very crude measure of the local size of a space (size near a point). A curve that is almost space-filling can still have topological dimension one, even if it fills up most of the area of a region. A fractal has an integer topological dimension, but in terms of the amount of space it takes up, it behaves like a higher-dimensional space.

The Hausdorff dimension measures the local size of a space taking into account the distance between points, the metric. Consider the number N(r) of balls of radius at most r required to cover X completely. When r is very small, N(r) grows polynomially with 1/r. For a sufficiently well-behaved X, the Hausdorff dimension is the unique number d such that N(r) grows as 1/rd as r approaches zero. More precisely, this defines the box-counting dimension, which equals the Hausdorff dimension when the value d is a critical boundary between growth rates that are insufficient to cover the space, and growth rates that are overabundant.

The Hausdorff dimension measures the local size of a space taking into account the distance between points, the metric. Consider the number N(r) of balls of radius at most r required to cover X completely. When r is very small, N(r) grows polynomially with 1/r. For a sufficiently well-behaved X, the Hausdorff dimension is the unique number d such that N(r) grows as 1/rd as r approaches zero. More precisely, this defines the box-counting dimension, which equals the Hausdorff dimension when the value d is a critical boundary between growth rates that are insufficient to cover the space, and growth rates that are overabundant.

For shapes that are smooth, or shapes with a small number of corners, the shapes of traditional geometry and science, the Hausdorff dimension is an integer agreeing with the topological dimension. But Benoit Mandelbrot observed that fractals, sets with noninteger Hausdorff dimensions, are found everywhere in nature. He observed that the proper idealization of most rough shapes you see around you is not in terms of smooth idealized shapes, but in terms of fractal idealized shapes:

For shapes that are smooth, or shapes with a small number of corners, the shapes of traditional geometry and science, the Hausdorff dimension is an integer agreeing with the topological dimension. But Benoit Mandelbrot observed that fractals, sets with noninteger Hausdorff dimensions, are found everywhere in nature. He observed that the proper idealization of most rough shapes you see around you is not in terms of smooth idealized shapes, but in terms of fractal idealized shapes:

Clouds are not spheres, mountains are not cones, coastlines are not circles, and bark is not smooth, nor does lightning travel in a straight line.[5]

Clouds are not spheres, mountains are not cones, coastlines are not circles, and bark is not smooth, nor does lightning travel in a straight line.

For fractals that occur in nature, the Hausdorff and box-counting dimension coincide. The packing dimension is yet another similar notion which gives the same value for many shapes, but there are well documented exceptions where all these dimensions differ.

For fractals that occur in nature, the Hausdorff and box-counting dimension coincide. The packing dimension is yet another similar notion which gives the same value for many shapes, but there are well documented exceptions where all these dimensions differ.

## Formal definitions形式化定义

### Hausdorff content豪斯多夫集

Let X be a metric space. If SX and d ∈ [0, ∞), the d-dimensional unlimited Hausdorff content of S is defined by

Let X be a metric space. If S ⊂ X and d ∈ [0, ∞), the d-dimensional unlimited Hausdorff content of S is defined by

X是度量空间。若SXd ∈ [0, ∞) ，则 Sd维无限 豪斯多夫集定义为

$\displaystyle{ C_H^d(S):=\inf\Bigl\{\sum_i r_i^d:\text{ there is a cover of } S\text{ by balls with radii }r_i\gt 0\Bigr\}. }$

$\displaystyle{ C_H^d(S):=\inf\Bigl\{\sum_i r_i^d:\text{ there is a cover of } S\text{ by balls with radii }r_i\gt 0\Bigr\}. }$

$\displaystyle{ C_H^d(S):=\inf\Bigl\{\sum_i r_i^d:\text{ there is a cover of } S\text{ by balls with radii }r_i\gt 0\Bigr\}. }$

In other words, $\displaystyle{ C_H^d(S) }$ is the infimum of the set of numbers $\displaystyle{ \delta \geq 0 }$ such that there is some (indexed) collection of balls $\displaystyle{ \{B(x_i,r_i):i\in I\} }$ covering S with ri > 0 for each i ∈ I that satisfies $\displaystyle{ \sum_{i\in I} r_i^d\lt \delta }$. (Here, we use the standard convention that inf Ø = ∞.)

In other words, $\displaystyle{ C_H^d(S) }$ is the infimum of the set of numbers $\displaystyle{ \delta \geq 0 }$ such that there is some (indexed) collection of balls $\displaystyle{ \{B(x_i,r_i):i\in I\} }$ covering S with ri > 0 for each i ∈ I that satisfies $\displaystyle{ \sum_{i\in I} r_i^d\lt \delta }$. (Here, we use the standard convention that inf Ø = ∞.)

### Hausdorff measurement豪斯多夫分形测量

The Hausdorff outer measure is different from the unbounded Hausdorf content in that rather than considering all possible coverings of S, we see what happens when the sizes of the balls go to zero. This is for $\displaystyle{ d \geq 0 }$, we define the d-dimensional Hausdorff outer measure of S as

The Hausdorff outer measure is different from the unbounded Hausdorf content in that rather than considering all possible coverings of S, we see what happens when the sizes of the balls go to zero. This is for $\displaystyle{ d \geq 0 }$, we define the d-dimensional Hausdorff outer measure of S as

$\displaystyle{ \mathcal{H}^d(S):=\lim_{r \to 0} \inf\Bigl\{\sum_i r_i^d:\text{ there is a cover of } S\text{ by balls with radii } 0 \lt r_i \lt r\Bigr\}. }$

$\displaystyle{ \mathcal{H}^d(S):=\lim_{r \to 0} \inf\Bigl\{\sum_i r_i^d:\text{ there is a cover of } S\text{ by balls with radii } 0 \lt r_i \lt r\Bigr\}. }$

$\displaystyle{ \mathcal{H}^d(S):=\lim_{r \to 0} \inf\Bigl\{\sum_i r_i^d:\text{有一个面} S\text{ 球 } 0 \lt r_i \lt r\Bigr\}. }$

### Hausdorff dimension豪斯多夫维数

The Hausdorff dimension of X is defined by

The Hausdorff dimension of X is defined by

X 的豪斯多夫维数定义为

$\displaystyle{ \dim_{\operatorname{H}}(X):=\inf\{d\ge 0: \mathcal{H}^d(X)=0\}. }$

$\displaystyle{ \dim_{\operatorname{H}}(X):=\inf\{d\ge 0: \mathcal{H}^d(X)=0\}. }$

Equivalently, dimH(X) may be defined as the infimum of the set of d ∈ [0, ∞) such that the d-dimensional Hausdorff measure of X is zero. This is the same as the supremum of the set of d ∈ [0, ∞) such that the d-dimensional Hausdorff measure of X is infinite (except that when this latter set of numbers d is empty the Hausdorff dimension is zero).

Equivalently, dimH(X) may be defined as the infimum of the set of d ∈ [0, ∞) such that the d-dimensional Hausdorff measure of X is zero. This is the same as the supremum of the set of d ∈ [0, ∞) such that the d-dimensional Hausdorff measure of X is infinite (except that when this latter set of numbers d is empty the Hausdorff dimension is zero).

## Examples实例

Dimension of a further fractal example. The Sierpinski triangle, an object with Hausdorff dimension of log(3)/log(2)≈1.58.[4]

• Countable sets have Hausdorff dimension 0.[6]
• 可数集拥有豪斯多夫维数0.[6]
• The Euclidean spacen has Hausdorff dimension n, and the circle S1 has Hausdorff dimension 1.[6]
• 教育空间n 有豪斯多夫维数 n，循环S1 拥有豪斯多夫维数1.[6]
• Fractals often are spaces whose Hausdorff dimension strictly exceeds the topological dimension.[5] For example, the Cantor set, a zero-dimensional topological space, is a union of two copies of itself, each copy shrunk by a factor 1/3; hence, it can be shown that its Hausdorff dimension is ln(2)/ln(3) ≈ 0.63.[7] The Sierpinski triangle is a union of three copies of itself, each copy shrunk by a factor of 1/2; this yields a Hausdorff dimension of ln(3)/ln(2) ≈ 1.58.[1] These Hausdorff dimensions are related to the "critical exponent" of the Master theorem for solving recurrence relations in the analysis of algorithms.
• 分形一般是那些豪斯多夫维数直接超过其拓扑维数的空间。例如康托尔集是一个o维拓扑空间，由两个自己复制而成，每一个复制品都是原来的三分之一，因此它的豪斯多夫维数是 ln(2)/ln(3) ≈ 0.63。[8]一个谢尔宾斯基三角是他自身三个复制的组合。每一个是原来的 1/2，它的豪斯多夫维数ln(3)/ln(2) ≈ 1.58。[1]在递归算法中解决递归关系时，这些豪斯多夫维数与算法分析主定理的临界指标相联系。
• Space-filling curves like the Peano curve have the same Hausdorff dimension as the space they fill.
• 空间填充曲线拥有和他们填充空间同样的豪斯多夫维数，如皮亚诺曲线
• The trajectory of Brownian motion in dimension 2 and above is conjectured to be Hausdorff dimension 2.[9]

Estimating the Hausdorff dimension of the coast of Great Britain

coast of Great Britain]]

[[大不列颠海岸

• 刘易斯·弗赖伊·理查森Lewis Fry Richardson has performed detailed experiments to measure the approximate Hausdorff dimension for various coastlines. His results have varied from 1.02 for the coastline of South Africa to 1.25 for the west coast of Great Britain.[5]
• 刘易斯·弗赖伊·理查森（Lewis Fry Richardson）已经通过豪斯多夫维数去测量了很多海岸线。它的结果涵盖从1.02的南非海岸线到1.25的大英帝国西海岸模型。]]

## Properties of Hausdorff dimension豪斯多夫维数特性

### Hausdorff dimension and inductive dimension 豪斯多夫维数和归纳维数

Let X be an arbitrary separable metric space. There is a topological notion of inductive dimension for X which is defined recursively. It is always an integer (or +∞) and is denoted dimind(X).

Let X be an arbitrary separable metric space. There is a topological notion of inductive dimension for X which is defined recursively. It is always an integer (or +∞) and is denoted dimind(X).

X任意可分度量空间。对于X 有一个递归定义的归纳维数拓扑概念。它总是一个整数(或 + ∞) ，记为dimind(X)。

Theorem. Suppose X is non-empty. Then

Theorem. Suppose X is non-empty. Then

$\displaystyle{ \dim_{\mathrm{Haus}}(X) \geq \dim_{\operatorname{ind}}(X). }$

$\displaystyle{ \dim_{\mathrm{Haus}}(X) \geq \dim_{\operatorname{ind}}(X). }$

Moreover,

Moreover,

$\displaystyle{ \inf_Y \dim_{\operatorname{Haus}}(Y) =\dim_{\operatorname{ind}}(X), }$

$\displaystyle{ \inf_Y \dim_{\operatorname{Haus}}(Y) =\dim_{\operatorname{ind}}(X), }$

where Y ranges over metric spaces homeomorphic to X. In other words, X and Y have the same underlying set of points and the metric dY of Y is topologically equivalent to dX.

where Y ranges over metric spaces homeomorphic to X. In other words, X and Y have the same underlying set of points and the metric dY of Y is topologically equivalent to dX.

These results were originally established by Edward Szpilrajn (1907–1976), e.g., see Hurewicz and Wallman, Chapter VII.模板:Full citation needed

These results were originally established by Edward Szpilrajn (1907–1976), e.g., see Hurewicz and Wallman, Chapter VII.

### Hausdorff dimension and Minkowski dimension 豪斯多夫维数和闵可夫斯基维度

The Minkowski dimension is similar to, and at least as large as, the Hausdorff dimension, and they are equal in many situations. However, the set of rational points in [0, 1] has Hausdorff dimension zero and Minkowski dimension one. There are also compact sets for which the Minkowski dimension is strictly larger than the Hausdorff dimension.

The Minkowski dimension is similar to, and at least as large as, the Hausdorff dimension, and they are equal in many situations. However, the set of rational points in [0, 1] has Hausdorff dimension zero and Minkowski dimension one. There are also compact sets for which the Minkowski dimension is strictly larger than the Hausdorff dimension.

### Hausdorff dimensions and Frostman measures 豪斯多夫维度和弗洛斯曼测度

If there is a measure μ defined on Borel subsets of a metric space X such that μ(X) > 0 and μ(B(x, r)) ≤ rs holds for some constant s > 0 and for every ball B(x, r) in X, then dimHaus(X) ≥ s. A partial converse is provided by Frostman's lemma.[citation needed][10]

If there is a measure μ defined on Borel subsets of a metric space X such that μ(X) > 0 and μ(B(x, r)) ≤ rs holds for some constant s > 0 and for every ball B(x, r) in X, then dimHaus(X) ≥ s. A partial converse is provided by Frostman's lemma.

### Behaviour under unions and products

If $\displaystyle{ X=\bigcup_{i\in I}X_i }$ is a finite or countable union, then

If $\displaystyle{ X=\bigcup_{i\in I}X_i }$ is a finite or countable union, then

$\displaystyle{ \dim_{\operatorname{Haus}}(X) =\sup_{i\in I} \dim_{\operatorname{Haus}}(X_i). }$

$\displaystyle{ \dim_{\operatorname{Haus}}(X) =\sup_{i\in I} \dim_{\operatorname{Haus}}(X_i). }$

This can be verified directly from the definition.

This can be verified directly from the definition.

If X and Y are non-empty metric spaces, then the Hausdorff dimension of their product satisfies[11]

If X and Y are non-empty metric spaces, then the Hausdorff dimension of their product satisfies

$\displaystyle{ \dim_{\operatorname{Haus}}(X\times Y)\ge \dim_{\operatorname{Haus}}(X)+ \dim_{\operatorname{Haus}}(Y). }$

$\displaystyle{ \dim_{\operatorname{Haus}}(X\times Y)\ge \dim_{\operatorname{Haus}}(X)+ \dim_{\operatorname{Haus}}(Y). }$

This inequality can be strict. It is possible to find two sets of dimension 0 whose product has dimension 1.[12] In the opposite direction, it is known that when X and Y are Borel subsets of Rn, the Hausdorff dimension of X × Y is bounded from above by the Hausdorff dimension of X plus the upper packing dimension of Y. These facts are discussed in Mattila (1995).

This inequality can be strict. It is possible to find two sets of dimension 0 whose product has dimension 1. In the opposite direction, it is known that when X and Y are Borel subsets of Rn, the Hausdorff dimension of X × Y is bounded from above by the Hausdorff dimension of X plus the upper packing dimension of Y. These facts are discussed in Mattila (1995).

## Self-similar sets自相似集合

Many sets defined by a self-similarity condition have dimensions which can be determined explicitly. Roughly, a set E is self-similar if it is the fixed point of a set-valued transformation ψ, that is ψ(E) = E, although the exact definition is given below.

Many sets defined by a self-similarity condition have dimensions which can be determined explicitly. Roughly, a set E is self-similar if it is the fixed point of a set-valued transformation ψ, that is ψ(E) = E, although the exact definition is given below.

Theorem. Suppose

Theorem. Suppose

$\displaystyle{ \psi_i: \mathbf{R}^n \rightarrow \mathbf{R}^n, \quad i=1, \ldots , m }$

$\displaystyle{ \psi_i: \mathbf{R}^n \rightarrow \mathbf{R}^n, \quad i=1, \ldots , m }$

are contractive mappings on Rn with contraction constant rj < 1. Then there is a unique non-empty compact set A such that

are contractive mappings on Rn with contraction constant rj < 1. Then there is a unique non-empty compact set A such that

Rn上的压缩常数rj'< 1的压缩映射。则有一个唯一的非空紧集A

$\displaystyle{ A = \bigcup_{i=1}^m \psi_i (A). }$

$\displaystyle{ A = \bigcup_{i=1}^m \psi_i (A). }$

The theorem follows from Stefan Banach's contractive mapping fixed point theorem applied to the complete metric space of non-empty compact subsets of Rn with the Hausdorff distance.[13]

The theorem follows from Stefan Banach's contractive mapping fixed point theorem applied to the complete metric space of non-empty compact subsets of Rn with the Hausdorff distance.

### The open set condition开集条件

To determine the dimension of the self-similar set A (in certain cases), we need a technical condition called the open set condition (OSC) on the sequence of contractions ψi.

To determine the dimension of the self-similar set A (in certain cases), we need a technical condition called the open set condition (OSC) on the sequence of contractions ψi.

There is a relatively compact open set V such that

There is a relatively compact open set V such that

$\displaystyle{ \bigcup_{i=1}^m\psi_i (V) \subseteq V, }$

$\displaystyle{ \bigcup_{i=1}^m\psi_i (V) \subseteq V, }$

where the sets in union on the left are pairwise disjoint.

where the sets in union on the left are pairwise disjoint.

The open set condition is a separation condition that ensures the images ψi(V) do not overlap "too much".

The open set condition is a separation condition that ensures the images ψi(V) do not overlap "too much".

Theorem. Suppose the open set condition holds and each ψi is a similitude, that is a composition of an isometry and a dilation around some point. Then the unique fixed point of ψ is a set whose Hausdorff dimension is s where s is the unique solution of[14]

Theorem. Suppose the open set condition holds and each ψi is a similitude, that is a composition of an isometry and a dilation around some point. Then the unique fixed point of ψ is a set whose Hausdorff dimension is s where s is the unique solution of

$\displaystyle{ \sum_{i=1}^m r_i^s = 1. }$

$\displaystyle{ \sum_{i=1}^m r_i^s = 1. }$

The contraction coefficient of a similitude is the magnitude of the dilation.

The contraction coefficient of a similitude is the magnitude of the dilation.

We can use this theorem to compute the Hausdorff dimension of the Sierpinski triangle (or sometimes called Sierpinski gasket). Consider three non-collinear points a1, a2, a3 in the plane R2 and let ψi be the dilation of ratio 1/2 around ai. The unique non-empty fixed point of the corresponding mapping ψ is a Sierpinski gasket and the dimension s is the unique solution of

We can use this theorem to compute the Hausdorff dimension of the Sierpinski triangle (or sometimes called Sierpinski gasket). Consider three non-collinear points a1, a2, a3 in the plane R2 and let ψi be the dilation of ratio 1/2 around ai. The unique non-empty fixed point of the corresponding mapping ψ is a Sierpinski gasket and the dimension s is the unique solution of

$\displaystyle{ \left(\frac{1}{2}\right)^s+\left(\frac{1}{2}\right)^s+\left(\frac{1}{2}\right)^s = 3 \left(\frac{1}{2}\right)^s =1. }$

$\displaystyle{ \left(\frac{1}{2}\right)^s+\left(\frac{1}{2}\right)^s+\left(\frac{1}{2}\right)^s = 3 \left(\frac{1}{2}\right)^s =1. }$

Taking natural logarithms of both sides of the above equation, we can solve for s, that is: s = ln(3)/ln(2). The Sierpinski gasket is self-similar and satisfies the OSC. In general a set E which is a fixed point of a mapping

Taking natural logarithms of both sides of the above equation, we can solve for s, that is: s = ln(3)/ln(2). The Sierpinski gasket is self-similar and satisfies the OSC. In general a set E which is a fixed point of a mapping

$\displaystyle{ A \mapsto \psi(A) = \bigcup_{i=1}^m \psi_i(A) }$
$\displaystyle{ A \mapsto \psi(A) = \bigcup_{i=1}^m \psi_i(A) }$


is self-similar if and only if the intersections

is self-similar if and only if the intersections

$\displaystyle{ H^s\left(\psi_i(E) \cap \psi_j(E)\right) =0, }$

$\displaystyle{ H^s\left(\psi_i(E) \cap \psi_j(E)\right) =0, }$

where s is the Hausdorff dimension of E and Hs denotes Hausdorff measure. This is clear in the case of the Sierpinski gasket (the intersections are just points), but is also true more generally:

where s is the Hausdorff dimension of E and Hs denotes Hausdorff measure. This is clear in the case of the Sierpinski gasket (the intersections are just points), but is also true more generally:

Theorem. Under the same conditions as the previous theorem, the unique fixed point of ψ is self-similar.

Theorem. Under the same conditions as the previous theorem, the unique fixed point of ψ is self-similar.

• Assouad dimension Assouad维数, another variation of fractal dimension that, like Hausdorff dimension, is defined using coverings by balls其他被球覆盖方式定义的分形维数，例如豪斯多夫维数。

## References参考文献

1. MacGregor Campbell, 2013, "5.6 Scaling and the Hausdorff Dimension," at Annenberg Learner:MATHematics illuminated, see [1], accessed 5 March 2015. 引用错误：无效<ref>标签；name属性“CampbellAnnenberg15”使用不同内容定义了多次
2. Gneiting, Tilmann; Ševčíková, Hana; Percival, Donald B. (2012). "Estimators of Fractal Dimension: Assessing the Roughness of Time Series and Spatial Data". Statistical Science. 27 (2): 247–277. arXiv:1101.1444. doi:10.1214/11-STS370.
3. Larry Riddle, 2014, "Classic Iterated Function Systems: Koch Snowflake", Agnes Scott College e-Academy (online), see [2], accessed 5 March 2015.
4. Keith Clayton, 1996, "Fractals and the Fractal Dimension," Basic Concepts in Nonlinear Dynamics and Chaos (workshop), Society for Chaos Theory in Psychology and the Life Sciences annual meeting, June 28, 1996, Berkeley, California, see [3], accessed 5 March 2015.
5. Mandelbrot, Benoît (1982). The Fractal Geometry of Nature. Lecture notes in mathematics 1358. W. H. Freeman. ISBN 0-7167-1186-9.
6. Schleicher, Dierk (June 2007). "Hausdorff Dimension, Its Properties, and Its Surprises". The American Mathematical Monthly (in English). 114 (6): 509–528. arXiv:math/0505099. doi:10.1080/00029890.2007.11920440. ISSN 0002-9890. 引用错误：无效<ref>标签；name属性“schleicher”使用不同内容定义了多次
7. Falconer, Kenneth (2003). Fractal Geometry: Mathematical Foundations and Applications (2nd ed.). John Wiley and Sons.
8. Falconer, Kenneth (2003). Fractal Geometry: Mathematical Foundations and Applications (2nd ed.). John Wiley and Sons.
9. Morters, Peres (2010). Brownian Motion. Cambridge University Press.
10. This Wikipedia article also discusses further useful characterizations of the Hausdorff dimension.模板:Clarify
11. Marstrand, J. M. (1954). "The dimension of Cartesian product sets". Proc. Cambridge Philos. Soc. 50 (3): 198–202. Bibcode:1954PCPS...50..198M. doi:10.1017/S0305004100029236.
12. Falconer, Kenneth J. (2003). Fractal geometry. Mathematical foundations and applications. John Wiley & Sons, Inc., Hoboken, New Jersey.
13. Falconer, K. J. (1985). "Theorem 8.3". The Geometry of Fractal Sets. Cambridge, UK: Cambridge University Press. ISBN 0-521-25694-1.
14. Hutchinson, John E. (1981). "Fractals and self similarity". Indiana Univ. Math. J. 30 (5): 713–747. doi:10.1512/iumj.1981.30.30055.
15. Hutchinson, John E. (1981). "Fractals and self similarity". Indiana Univ. Math. J. 30 (5): 713–747. doi:10.1512/iumj.1981.30.30055.

• "La dimension et la mesure". Fundamenta Mathematicae. 28: 81–9. 1937. Unknown parameter |author作者= ignored (help)
" ignored (help); Text " 第198-202页

" ignored (help); Text " 第三期

" ignored (help)

| doi=10.1017/S0305004100029236|bibcode = 1954PCPS...50..198M }}


10.1017 / S0305004100029236 | bibcode 1954PCPS... 50. . 198 m }

| isbn=978-0-521-65595-8 | year=1995}}


| isbn 978-0-521-65595-8 | year 1995}

• Falconer, Kenneth (2003). Fractal Geometry: Mathematical Foundations and Applications (2nd ed.). John Wiley and Sons.