Fractals · Geometric series · Complex iteration
How can a shape have zero area, infinite total hole-boundary length, and a dimension between 1 and 2? Start with the Sierpiński carpet, then test which complex numbers belong to the Mandelbrot set.
Watch: fractals, dimension, and the Mandelbrot set
Your full lesson continues below the video. Explore the 10-slide carousel, follow the worked video examples, and use the formula sheet and practice tasks to make the ideas your own.
One repeated rule
Keep eight of nine squares, then repeat the construction inside each remaining square.
Follow the construction →A dimension of 1.8928…
Count eight copies at one-third scale. Logarithms turn the scaling rule into a dimension.
Work the calculation →Bounded or escaping?
Iterate z² + c from zero. An escaping critical orbit excludes c from the Mandelbrot set.
See the escape proof →What Is a Fractal?
A fractal is a mathematical set or geometric object with meaningful structure across scales. Magnifying it continues to reveal organized detail. Classical fractals often display exact self-similarity and may have a non-integer dimension, while other fractals—such as the Mandelbrot set—have more complicated, nonrepeating local structure.
There is no single elementary checklist that captures every object called a fractal. Exact self-similarity is common but not universal. Non-integer dimension is common but not required. The most reliable unifying idea is that ordinary Euclidean geometry does not adequately describe how the object’s detail behaves as scale changes.
Complexity does not require a complicated rule. Repetition, scaling, and feedback can turn a short instruction into geometry with inexhaustible mathematical depth.
The full visual lesson · 10 slides
See the structure at every scale
Swipe sideways or use the numbered links. Select a slide to jump to its full-size image and explanation. Keyboard users can focus the carousel and use the arrow keys.
Worked recap of the video
The Sierpiński carpet: area, edge length, and dimension
Begin with a unit square. Divide it into nine equal squares and remove the open center square. Repeat inside each of the eight remaining squares.
1. Count first, then compute the area
After \(n\) steps, there are \(8^n\) squares, each with side length \(3^{-n}\). Therefore,
The removed areas form a geometric series with first term \(1/9\) and ratio \(8/9\):
2. Add the boundaries of the holes
At step \(k\), there are \(8^{k-1}\) new holes, each with perimeter \(4\cdot3^{-k}\). Their combined new boundary length is
These positive terms grow rather than approach zero, so their sum diverges. The carpet has zero area, while the combined lengths of its hole boundaries are infinite. The Koch snowflake example later in the written lesson illustrates a different construction with finite positive area and infinite perimeter.
3. Derive the dimension
The standard carpet consists of eight self-similar copies at scale \(1/3\). Its similarity dimension satisfies
For this standard construction, the pieces satisfy the separation condition needed for this similarity dimension to equal the Hausdorff dimension. The formula is not a shortcut for every object called a fractal.
4. Test the video’s two Mandelbrot parameters
Starting with \(z_0=0\), apply \(z_{n+1}=z_n^2+c\):
| Parameter | Orbit from zero | Conclusion |
|---|---|---|
| \(c=2\) | \(0\to2\to6\to38\to\cdots\) | The modulus exceeds 2 at 6, so the orbit escapes and \(2\notin M\). |
| \(c=i\) | \(0\to i\to-1+i\to-i\to-1+i\to\cdots\) | The orbit enters a two-cycle. It stays bounded, so \(i\in M\). |
Check the complex arithmetic: \((-1+i)^2+i=-2i+i=-i\), and \((-i)^2+i=-1+i\). Returning to a previous value makes the same cycle repeat.
Try the written examples and read the escape-radius proof ↓ · Back to the video ↑
Explore the complete fractals lesson
How simple rules create infinite mathematical structure.
Fractals are not merely psychedelic pictures or objects that “look complicated.” They are mathematical structures organized across scales. Some are built by exact self-similarity, some emerge from repeated iteration, some have non-integer dimension, and some form the boundaries of chaotic dynamical systems. This lesson develops the mathematics from the ground up: definition, iteration, the Mandelbrot set, escape orbits, Julia sets, fractal dimension, chaos, natural patterns, myths, worked examples, and mastery tasks.
Reading level: layered for curious beginners, high-school and undergraduate students, and advanced readers who want the precise mathematics.
What Makes a Fractal a Fractal?
The word fractal is used for several related kinds of mathematical objects. The following properties are common, but no single one is a complete universal definition.
- Structure across scales: magnification reveals additional organized detail rather than a featureless smooth object.
- Self-similarity: smaller portions may resemble the whole. The resemblance can be exact, approximate, or statistical.
- Iterative construction: many fractals arise by applying the same transformation repeatedly.
- Unusual dimension: a fractal may occupy space in a way that is more substantial than a line but less complete than a surface.
- Fine boundary structure: some fractals appear as boundaries separating different dynamical behaviors.
Exact, Approximate, and Statistical Self-Similarity
| Type | Meaning | Example |
|---|---|---|
| Exact self-similarity | Scaled pieces reproduce the whole exactly. | Cantor set, Koch curve, Sierpiński triangle |
| Approximate self-similarity | Smaller pieces resemble the whole but are not exact copies. | Many branching biological structures |
| Statistical self-similarity | Scaling laws persist statistically rather than point by point. | Idealized models of coastlines, clouds, and rough surfaces |
| Complex multiscale structure | New local structure appears at each scale without exact global repetition. | Mandelbrot set |
The Mandelbrot set is often casually called self-similar, but it is not globally exactly self-similar like the Cantor set. Its boundary contains repeated and near-repeated structures, including miniature Mandelbrot-like copies, while different locations reveal genuinely different geometry.
A Short History of Fractal Geometry
Fractal geometry did not appear in one moment. It grew from several mathematical objects that challenged the smooth geometry of points, lines, curves, and surfaces.
Georg Cantor: the middle-thirds Cantor set showed that an uncountable set can have total length zero and can reproduce itself at smaller scales.
Helge von Koch: the Koch curve supplied an elementary geometric example of a continuous curve with no tangent, infinite length, and exact self-similarity.
Wacław Sierpiński: recursive constructions such as the Sierpiński triangle revealed systematic holes and non-integer scaling behavior.
Gaston Julia and Pierre Fatou: early twentieth-century work on iterated complex functions established the foundations of what is now called complex dynamics.
Benoît Mandelbrot: in the 1970s and early 1980s, Mandelbrot helped unite these previously isolated examples under the language of fractal geometry and used emerging computer graphics to reveal the extraordinary structures produced by iteration.
See the MacTutor History of Fractal Geometry for a documented historical overview.
Iteration: The Engine Behind Fractals
An iteration feeds the output of one step back into the same rule. If the transformation is denoted by \(T\), then a sequence of stages can be written as
In dynamical systems, the resulting sequence
is called the orbit of the initial point. An orbit may settle to a fixed point, enter a periodic cycle, remain bounded without repeating, or escape without bound.
Iteration Is Not the Same as Recursion
The words are closely related but emphasize different viewpoints. A recursive definition describes an object in terms of earlier stages. Iteration emphasizes repeatedly applying a transformation. Many fractal constructions can be described both ways.
Example: The Sierpiński Triangle
Begin with a filled equilateral triangle.
Divide it into four congruent triangles.
Remove the central triangle.
Repeat the same operation on every remaining triangle.
After each step, three copies remain, each scaled by a factor of one half. That scaling information will later determine the Sierpiński triangle’s dimension.
What Is the Mandelbrot Set?
For each complex number \(c\), define the quadratic function
Start at the critical point \(z_0=0\), then iterate:
The parameter \(c\) belongs to the Mandelbrot set when the orbit \((z_n)\) remains bounded.
The Mandelbrot set is not one orbit. It is a map of an entire parameter plane. Every point \(c\) defines a different quadratic dynamical system, and the picture records which of those systems keep the critical orbit bounded.
How a Computer Draws the Mandelbrot Set
A computer assigns one complex parameter \(c=a+bi\) to each pixel, runs the iteration, and checks whether the orbit escapes. Outside points are often colored according to how quickly they escape. Points that have not escaped after a chosen iteration limit are drawn as if they belong to the set, producing a numerical approximation.
for each pixel:
convert pixel position to c = a + bi
z = 0
repeat up to MAX_ITERATIONS:
z = z*z + c
if abs(z) > 2:
color by escape time
stop
if no escape was detected:
color as an unresolved bounded pointFailing to escape after finitely many steps does not by itself prove that a point belongs to the Mandelbrot set. A rendered image is a finite numerical approximation to an infinite mathematical definition.
How the Mandelbrot Escape Test Works
Example A: \(c=0\)
The orbit is bounded, so
Example B: \(c=-1\)
The orbit enters a period-two cycle, so
Example C: \(c=1\)
The orbit escapes to infinity, so
Why the Escape Radius Is 2
First suppose \(|c|\le2\). If an iterate satisfies \(|z_n|>2\), then the reverse triangle inequality gives
Subtracting 2 from both sides yields
Therefore, for every integer \(k\ge1\),
The right-hand side tends to infinity, so \(|z_{n+k}|\to\infty\). This proves escape when \(|c|\le2\).
Now suppose \(|c|>2\). Since \(z_1=c\),
Whenever \(|z_n|\ge|c|>2\),
Because \(|c|-1>1\), the magnitudes grow at least geometrically. Hence this case also escapes to infinity.
Choose a complex parameter \(c\).
Set \(z_0=0\).
Repeat \(z_{n+1}=z_n^2+c\).
If an iterate satisfies \(|z_n|>2\), stop: the orbit will escape and \(c\notin M\).
If no escape is detected after the chosen iteration limit, mark the pixel as unresolved or provisionally bounded. More iterations improve the approximation near the boundary.
Mandelbrot Set vs. Filled Julia Set vs. Julia Set
The Mandelbrot set and Julia sets are built from the same quadratic family, but they answer different questions.
The Mandelbrot Set Varies the Parameter
In the Mandelbrot set, we vary the parameter \(c\), always begin with the critical point \(z_0=0\), and ask whether that one critical orbit remains bounded.
The Filled Julia Set Fixes \(c\) and Varies the Initial Point
Fix one parameter \(c\). The filled Julia set \(K_c\) consists of all initial points whose orbits remain bounded:
The Julia Set Is the Boundary
The Julia set is not the entire bounded-orbit region. It is the boundary of the filled Julia set:
This boundary separates qualitatively different orbit behavior and is where the most intricate complex dynamics occur.
| Object | What Varies? | Question | Plane |
|---|---|---|---|
| Mandelbrot set \(M\) | Parameter \(c\) | Does the critical orbit of zero remain bounded? | Parameter plane |
| Filled Julia set \(K_c\) | Initial point \(z_0\), with \(c\) fixed | Which initial orbits remain bounded? | Dynamical plane |
| Julia set \(J_c\) | Boundary points, with \(c\) fixed | Where is the boundary between dynamical behaviors? | Dynamical plane |
For the quadratic family \(f_c(z)=z^2+c\), the Julia set is connected exactly when the critical parameter belongs to the Mandelbrot set:
See Julia Sets and the Mandelbrot Set for a concise mathematical introduction to bounded orbits, filled Julia sets, Julia sets, connectedness, and the Mandelbrot criterion.
What Is Fractal Dimension?
Ordinary Euclidean dimension counts independent directions:
Fractal dimension asks a different scaling question: how rapidly does the amount of visible structure grow as the measuring scale becomes smaller?
The Similarity-Dimension Formula
Suppose an exactly self-similar object consists of \(N\) congruent copies, each scaled by a factor \(r\), where \(0<r<1\). Under the usual separation assumptions, its similarity dimension is
Koch Curve
The Koch curve contains four self-similar copies, each scaled by one third:
The curve is more space-filling than an ordinary line but does not fill a plane:
Sierpiński Triangle
The Sierpiński triangle contains three copies, each scaled by one half:
The formula \(D=\log N/\log(1/r)\) applies directly to equal-scale self-similar objects under suitable separation conditions. More general fractals require broader notions such as box-counting or Hausdorff dimension.
How Fractals Connect to Chaos Theory
A fractal is a geometric object or set. Chaos describes long-term dynamical behavior, especially deterministic systems with sensitive dependence on initial conditions. The subjects overlap because chaotic systems often produce fractal invariant sets, basin boundaries, and strange attractors.
Discrete Complex Dynamics
The rule is deterministic, but orbit behavior can change dramatically as the parameter or initial point changes. Julia sets form fractal boundaries between different dynamical regions.
Continuous Dynamics: The Lorenz System
The Lorenz equations are
For classical parameter choices, trajectories can approach the Lorenz strange attractor. The evolution is deterministic, yet nearby initial states separate rapidly, and the attractor has fractal geometric structure.
Chaos describes how a system evolves; fractals can describe the geometry organized by that evolution.
Continue the Chaos Theory Cluster
Deriving Fractal Dimension From Self-Similarity
Why should the dimension formula be \(D=\log N/\log(1/r)\)? The derivation comes from the scaling law for ordinary dimension.
Start With Ordinary Scaling
If an object has dimension \(D\) and is scaled by a linear factor \(r\), then its \(D\)-dimensional size scales by \(r^D\). If \(N\) scaled copies exactly rebuild the original object, their total scaled size must equal one original object:
Now solve for \(D\):
Trust Check: Does the Formula Recover a Square?
Divide a square into four smaller squares, each scaled by one half. Then
The scaling formula correctly reproduces the ordinary dimension of a square.
Cantor Set: Dimension Below 1
The middle-thirds Cantor set contains two self-similar copies, each scaled by one third:
It is more substantial than a finite or countable collection of isolated points, yet it occupies less of the line than a full one-dimensional interval.
Koch Curve: Dimension Above 1
The Koch curve is a curve, but its detail grows too quickly to behave like a smooth one-dimensional object.
For the full real-analysis story—uncountability, zero length, perfectness, nowhere density, and self-similarity—read The Cantor Set: Infinite Points, Zero Length, and Real Analysis .
Box-Counting Dimension and Hausdorff Dimension
Similarity dimension is elegant but specialized. Box-counting and Hausdorff dimension apply to much broader classes of sets.
Box-Counting Dimension
Cover a set \(E\) with a grid of boxes of side length \(\varepsilon\). Let \(N(\varepsilon)\) be the number of boxes needed to intersect the set. If the limit exists, the box-counting dimension is
If the ordinary limit fails to exist, mathematicians use upper and lower box-counting dimensions defined with limsup and liminf.
Hausdorff Dimension
Hausdorff dimension permits coverings by sets of varying sizes rather than one fixed grid. For each exponent \(s\), it studies sums of the form
The Hausdorff dimension is the critical value of \(s\) where the associated Hausdorff measure changes from infinity to zero. This definition is technically deeper, but it is powerful enough to distinguish very irregular sets.
A Deep Mandelbrot Fact
The Mandelbrot set contains interior, so the set itself has Hausdorff dimension two. More surprisingly, Shishikura proved that its boundary also has Hausdorff dimension two:
A set can therefore have the full Hausdorff dimension of the plane while still being a boundary with extraordinarily intricate local structure.
See Mitsuhiro Shishikura, The Hausdorff Dimension of the Boundary of the Mandelbrot Set and Julia Sets .
How Can a Fractal Have Infinite Perimeter but Finite Area?
Begin with an equilateral triangle of side length one. At every stage of the Koch snowflake, replace each segment with four segments, each one third as long.
Perimeter
The stage-\(n\) perimeter is
Since \(4/3>1\),
Area
Let \(A_0=\sqrt3/4\) be the area of the original triangle. The total added area is a convergent geometric series:
The Koch snowflake encloses a finite area even though its boundary has infinite length. Fractal geometry separates ideas that ordinary smooth geometry often makes us expect to move together.
Are Fractals Really Everywhere in Nature?
Nature often displays branching, roughness, and scale-dependent patterns that are well modeled with fractal ideas. But physical objects are not ideal mathematical fractals extending through infinitely many scales.
| Natural Structure | What the Fractal Model Captures | Physical Limitation |
|---|---|---|
| Coastlines | Measured length depends on the measuring scale; roughness can follow scaling laws. | Rock grains, erosion scales, maps, and measurement resolution impose cutoffs. |
| River networks | Branching and drainage organization across multiple scales. | Topography, geology, and finite watershed size limit the scaling range. |
| Trees and ferns | Repeated branching patterns and approximate self-similarity. | Cells, vascular constraints, gravity, and growth history prevent infinite repetition. |
| Lungs and blood vessels | Hierarchical branching used to distribute flow through space. | Biological tissues and vessel diameters have finite lower scales. |
| Lightning and clouds | Irregular branching or statistical roughness. | Atmospheric physics determines a finite range of observable scales. |
The Coastline Paradox
Measuring a coastline with a shorter ruler typically produces a larger measured length because the smaller ruler follows more bays, inlets, and irregularities. Mandelbrot’s famous 1967 paper used statistical self-similarity and fractional dimension to clarify this scale dependence.
Benoît Mandelbrot, How Long Is the Coast of Britain? Statistical Self-Similarity and Fractional Dimension .
Myth Check
| Claim | Verdict | Better Statement |
|---|---|---|
| Every complicated natural shape is a fractal. | False | Complex appearance alone does not establish a scaling law or fractal dimension. |
| Self-similarity must always be exact. | False | Natural systems often show approximate or statistical self-similarity. |
| Physical objects contain infinite detail. | False | Mathematical fractals can continue indefinitely; physical systems have lower and upper cutoffs. |
| Every spiral, branch, or rough boundary is fractal. | False | A fractal claim requires a defined scaling property, dimension, or dynamical construction. |
Mathematical fractals can continue forever. Physical nature eventually runs out of scale.
What Are Fractals Used For?
Fractal methods are useful when shape, roughness, branching, or complexity changes with scale. Applications do not require claiming that every object is a perfect fractal.
- Computer graphics and procedural generation: iterative and recursive rules create mountains, clouds, branching vegetation, textures, and virtual landscapes without specifying every detail by hand.
- Scientific image analysis: box-counting and related scaling measurements can quantify texture, branching, or boundary complexity in carefully defined datasets.
- Antenna engineering: self-similar geometries can produce compact multiband designs, although the engineering performance depends on the complete electromagnetic design rather than the word “fractal” alone.
- Geoscience and hydrology: scaling tools help analyze rough terrain, drainage networks, shorelines, and spatial clustering over finite ranges.
- Dynamical systems: fractal basin boundaries and strange attractors reveal how deterministic systems organize complicated long-term behavior.
- Mathematical analysis: Cantor-type sets, Hausdorff measure, nowhere differentiable curves, and complex dynamics test and extend the concepts of length, area, dimension, continuity, and boundary.
Why Fractals Matter
Fractals matter because they connect ideas that appear separate in an ordinary course sequence:
- Geometry: scaling, self-similarity, roughness, and dimension.
- Complex numbers: the dynamics of \(z^2+c\).
- Sequences: orbits generated by repeated iteration.
- Real analysis: limits, infinite constructions, measure, and Cantor sets.
- Differential equations: strange attractors and nonlinear dynamics.
- Computation: finite algorithms revealing the structure of infinite definitions.
Fractals show us that complexity can be born from simplicity. A short rule can generate an orbit, an orbit can organize a boundary, and a boundary can contain structure at every mathematically accessible scale.
Fractals Formula Sheet
General Iteration
Quadratic Complex Iteration
Mandelbrot Set
Escape Criterion
Filled Julia Set and Julia Set
Similarity Dimension
Box-Counting Dimension
Common Fractal Mistakes
- Calling every complicated image a fractal. Complexity of appearance is not enough; identify a scaling law, iterative construction, dimension, or dynamical boundary.
- Saying every fractal is exactly self-similar. Exact, approximate, statistical, and more general multiscale structures must be distinguished.
- Assuming every fractal has non-integer dimension. Some fractal sets have integer Hausdorff dimension; the boundary of the Mandelbrot set has dimension two.
- Confusing the parameter plane with the dynamical plane. The Mandelbrot set varies \(c\); a filled Julia set fixes \(c\) and varies \(z_0\).
- Calling the filled Julia set the Julia set. The bounded-orbit region is \(K_c\); the Julia set is the boundary \(J_c=\partial K_c\).
- Treating “did not escape after 500 iterations” as a proof of membership. Finite escape-time rendering is a numerical approximation.
- Using the similarity-dimension formula without checking its hypotheses. Equal-scale self-similarity and suitable separation are needed for the direct formula.
- Calling fractals and chaos synonyms. Fractals describe geometry; chaos describes dynamical behavior. They overlap but are not the same concept.
- Claiming nature contains literal infinite detail. Physical systems show finite-range, approximate, or statistical fractal behavior.
Prove That You Understand Fractals
Close the article before doing these. Retrieval is part of learning.
- Rewrite the Mandelbrot iteration \(z_{n+1}=z_n^2+c\) and the initial condition \(z_0=0\) from memory.
- Calculate the first five orbit values for \(c=0\), \(c=-1\), and \(c=1\), then classify each parameter.
- State the escape rule aloud and explain why crossing radius two forces the orbit to diverge.
- Explain the difference among the Mandelbrot set \(M\), filled Julia set \(K_c\), and Julia set \(J_c\) without looking back.
- Derive \(D=\log N/\log(1/r)\) starting from \(Nr^D=1\).
- Compute the similarity dimensions of the Cantor set, Koch curve, and Sierpiński triangle from memory.
- Say the distinction aloud until it is automatic: chaos is dynamics; fractals are geometry.
- Test a natural-fractal claim: choose one natural pattern and state exactly what data would be needed to test finite-range fractal scaling rather than merely asserting it.
Rewrite perfect solutions. Speak every transition. If you cannot explain why one line becomes the next, the mathematics is not yet yours.
Fractals FAQ
What is a fractal in simple terms?
A fractal is a mathematical set or shape with organized structure across scales. Magnifying it reveals additional detail, and many classical fractals are built from repeated scaled copies or iterative rules.
What are the main characteristics of a fractal?
Common characteristics include structure across scales, exact or approximate self-similarity, iterative construction, intricate boundaries, and a dimension that may differ from ordinary Euclidean dimension. Not every fractal has every one of these properties.
What is the Mandelbrot set?
The Mandelbrot set is the set of complex parameters c for which the orbit of zero under the iteration z squared plus c remains bounded.
What equation generates the Mandelbrot set?
Start with z zero equal to zero and repeatedly calculate z sub n plus one equal to z sub n squared plus c. The parameter c belongs to the Mandelbrot set when that orbit remains bounded.
Why is the Mandelbrot escape radius equal to 2?
Once an orbit value has modulus greater than two, the excess above radius two grows at least geometrically, forcing the orbit’s magnitudes to diverge.
Does a black pixel prove that a point is in the Mandelbrot set?
Not automatically. In a finite rendering, black often means only that the orbit did not escape before the chosen iteration limit. Boundary points may require much deeper computation or theoretical analysis.
What is the difference between the Mandelbrot set and a Julia set?
The Mandelbrot set varies the parameter c and tests the critical orbit starting at zero. A Julia set fixes c and describes the boundary of bounded behavior in the dynamical plane as the initial point varies.
What is the difference between a Julia set and a filled Julia set?
The filled Julia set K sub c contains all initial points with bounded orbits. The Julia set J sub c is the boundary of K sub c.
What is fractal dimension?
Fractal dimension measures how an object’s detail or space-filling behavior changes with scale. Depending on the context, mathematicians use similarity, box-counting, or Hausdorff dimension.
How can a dimension be fractional?
Dimension can be defined through scaling rather than only by counting directions. A self-similar object made of N copies scaled by r has similarity dimension log N divided by log of one over r, which need not be an integer.
What is the dimension of the Cantor set?
The middle-thirds Cantor set has similarity and Hausdorff dimension log 2 over log 3, approximately 0.63093.
What is the dimension of the Koch curve?
The Koch curve has dimension log 4 over log 3, approximately 1.26186, so its scaling behavior lies between that of a line and a surface.
Can a fractal have an integer dimension?
Yes. Non-integer dimension is common but not required. A deep theorem states that the boundary of the Mandelbrot set has Hausdorff dimension two.
Are fractals and chaos the same thing?
No. Fractals describe geometric structure, while chaos describes deterministic dynamical behavior with properties such as sensitive dependence. Chaotic systems often produce fractal attractors or boundaries, which is why the two subjects are closely connected.
Are fractals really found in nature?
Many natural structures show approximate or statistical fractal behavior over limited ranges of scale. Physical objects are not perfect infinite fractals because atoms, cells, material thickness, and finite size impose cutoffs.
Does the Mandelbrot set contain infinite detail?
As a mathematical set, its boundary has structure at arbitrarily small scales. A computer image shows only a finite-resolution approximation of that infinite mathematical object.
Who invented fractals?
No single person invented every fractal idea. Cantor, von Koch, Sierpiński, Julia, Fatou, and others created foundational examples and theories. Benoît Mandelbrot coined the term fractal and unified the field into modern fractal geometry.
What are fractals used for?
Fractal methods are used in procedural graphics, scaling analysis, scientific image analysis, roughness and branching models, dynamical systems, selected antenna designs, and mathematical studies of measure, dimension, and complex boundaries.
Fact-Check References
- MacTutor History of Mathematics: A History of Fractal Geometry
- Cornell Mathematics Notes: Julia Sets and the Mandelbrot Set
- Quadratic Maps: Mandelbrot and Julia Sets — Introduction to Fractal Geometry and Chaos
- Benoît Mandelbrot: Fractal Aspects of the Iteration of Complex Mappings
- Benoît Mandelbrot: How Long Is the Coast of Britain?
- Mitsuhiro Shishikura: Hausdorff Dimension of the Mandelbrot Boundary and Julia Sets
- Scholarpedia: Attractors and Strange Attractors