Fact-checked by Grok 2 weeks ago

Space-filling curve

A space-filling curve is a continuous surjective from the unit [0,1] onto the unit square [0,1] \times [0,1], representing a one-dimensional that densely fills a two-dimensional region by passing through every point in the square. These curves, also known as Peano curves, challenge intuitive notions of dimensionality, as they are topologically one-dimensional yet cover a two-dimensional space without gaps in the limit. The concept originated in the late 19th century when Italian mathematician constructed the first such curve in 1890, described in his paper "Sur une courbe qui remplit toute une aire plane," proving the existence of a mapping the interval to the square. In 1891, provided a more explicit and geometrically intuitive construction, now known as the , which avoids self-intersections at finite stages and has become one of the most studied examples. Subsequent developments include variants like the Sierpinski curve (1912) and the Lebesgue curve (1904), each offering different properties such as locality preservation or computational efficiency. Space-filling curves exhibit paradoxical properties: they are nowhere differentiable and have a of 2, matching the filled space, despite being parameterized by a one-dimensional . In higher dimensions, generalizations map [0,1] onto the unit [0,1]^n for n > 2, enabling similar filling behaviors. These mathematical objects have profound implications in , demonstrating that continuous images of low-dimensional sets can be high-dimensional. Beyond , space-filling curves find extensive applications in , particularly for multidimensional data indexing where they linearize higher-dimensional spaces while approximately preserving spatial locality, aiding in database queries, , and numerical simulations. In , they facilitate efficient domain decomposition by ordering data points to minimize communication overhead in scientific applications like finite element methods. Other uses include sensor network , where topology-dependent variants adapt to variations, and graphics rendering for .

Fundamentals

Definition

A space-filling curve is fundamentally understood within the framework of , where topological spaces provide the structure for defining and . of a between topological spaces means that the preimage of every in the codomain is open in the domain, ensuring no abrupt jumps in the mapping. , a key property here, means that every open cover of the space has a finite subcover; notably, the closed [0,1] equipped with the standard is compact, and the continuous image of a compact set is itself compact. Formally, a space-filling curve is defined as a continuous f: [0,1] \to [0,1]^n for some n \geq 2, where the image f([0,1]) coincides exactly with the entire unit cube [0,1]^n. Surjectivity guarantees that every point in the unit cube is attained by at least one value in [0,1], while preserves the connectedness of the , allowing the curve to weave through the higher-dimensional without discontinuities. This parameterization by a one-dimensional thus traces a that densely covers—and ultimately fills—the n-dimensional cube, often with multiplicities where multiple values map to the same point. The existence of such curves presents an intuitive , as they defy the everyday expectation that a one-dimensional object cannot completely occupy a higher-dimensional volume, challenging core notions of and even assumptions underlying results like the , which posits that simple closed curves in the plane separate the plane into interior and exterior regions without filling area. To grasp this counterintuitively, consider a simple non-rigorous approximation: a polygonal path that snakes back and forth across a square, much like parallel lines mowing a , starting at one side and covering horizontal strips; refining this by making the strips narrower and adding vertical connections at the ends allows the path to pass arbitrarily close to every point in the square, suggesting how can enable full coverage in the limit.

History

The concept of space-filling curves emerged from early investigations into the nature of infinity and dimension in . In 1883, demonstrated that the unit interval and the unit square possess the same , implying the existence of a between a one-dimensional and a two-dimensional one, though he emphasized that such mappings were not necessarily continuous. This work laid the intuitive foundation for space-filling curves by challenging classical notions of dimension, influencing the broader development of and . The first explicit construction of a continuous space-filling curve appeared in 1890, when published a surjective continuous map from the unit interval to the unit square in Mathematische Annalen. , now known as the , filled the entire square. In 1891, introduced a variant in another Mathematische Annalen article, designing a curve that was easier to visualize and iterate through recursive subdivision, which became widely adopted for its geometric appeal. Subsequent advancements explored pathological properties and generalizations. In 1903, William F. Osgood constructed Jordan curves of positive area that are nowhere differentiable, providing examples of pathological curves related to but distinct from space-filling curves. extended these constructions in 1915 to arbitrary finite dimensions, enabling surjective continuous mappings from the unit interval to the unit n-cube. These developments intertwined with evolving topological theories, including the Hahn–Mazurkiewicz theorem (proved around 1914), which characterizes precisely when a admits a space-filling curve parameterization. In 1932, Tadeusz Ważewski applied similar ideas to the infinite-dimensional , showing continuous images that fill the space.

Constructions

General Outline

Space-filling curves are typically constructed through an iterative process that generates a of increasingly refined approximations, ultimately converging to a continuous surjective from the unit [0,1] onto the unit square [0,1]^2. This abstract framework emphasizes refinement without specifying particular path orderings, focusing instead on the subdivision and mapping principles that ensure coverage in the limit. The process begins with an initial simple polygonal approximation, such as a straight line or basic polyline within the unit square, serving as the zeroth iterate f_0: [0,1] \to [0,1]^2. In each , the [0,1] is uniformly subdivided into m equal subintervals, where m \geq 2 is an , and the range square is partitioned into m^2 congruent subsquares, each of side length $1/m. The next approximation f_{n+1} is then defined by each subinterval affinely onto a scaled and translated copy of the previous f_n that traverses a corresponding subsquare, connecting them sequentially to maintain . This refinement scales the previous self-similarly into each subsquare while linking them sequentially to form a single connected path. The sequence \{f_n\} consists of continuous piecewise linear functions, with the number of linear segments in f_n growing as O((m^2)^n). Due to the compactness of [0,1] and [0,1]^2, the space of continuous functions C([0,1], [0,1]^2) equipped with the supremum is complete, ensuring that \{f_n\} converges uniformly to a limit f = \lim_{n \to \infty} f_n, which is continuous by the . The limit f is surjective onto [0,1]^2 because the images of the approximations cover denser and denser grids of points in the square, exhausting it in the limit via the density of the dyadic rationals or analogous grids. A key challenge in this construction lies in selecting a systematic ordering of the subsquares at each stage to guarantee surjectivity without introducing discontinuities or gaps; an improper ordering may fail to approach certain regions, preventing the limit from filling the space completely. This iterative principle underpins various space-filling curves, with Peano's 1890 construction providing an early realization of the approach.

Specific Examples

The Peano curve, first constructed by Giuseppe Peano in 1890, fills the unit square through an iterative process that divides the square into nine equal subsquares arranged in a 3×3 grid. At each iteration, the curve traverses these subsquares in a continuous path pattern, starting from the bottom-left subsquare and visiting all nine while maintaining continuity at the boundaries. The first-order approximation forms a simple polyline path; the second-order refines this by applying the pattern within each of the nine segments, creating 81 smaller path segments; and the third-order further subdivides into 729 segments, increasingly densely approximating the full square filling in the limit. The Lebesgue curve, constructed by in 1904, provides another early example using a subdivision similar to Peano's but with a mapping based on the . It represents points in [0,1] using ternary expansions avoiding the digit 1 (the Cantor set), interleaving even and odd digits to form binary coordinates for x and y, then extends linearly to fill the square, preserving some locality better than Peano's original. The , developed by in 1891 as a variant of Peano's construction, instead divides the unit square into four equal subsquares per iteration, promoting better preservation of spatial locality by keeping nearby points on the curve geometrically close. It builds through recursive U-shaped patterns: the zeroth-order is a single point; the first-order connects the four subsquares in a U , entering from the bottom-left and exiting from the top-right while rotating orientations (e.g., upward U in the bottom row, downward in the top); the second-order replaces each segment with the first-order pattern, yielding 16 subsquares traversed in a more intricate interlocking path; and the third-order expands to 64 subsquares, with the curve's fractal-like structure becoming evident as it fills denser regions without long jumps. Iteration rules can be pseudocode-like: define base orientations (N, S, E, W for north, south, east, west facing), then recursively compose four rotated/reflected copies connected at midpoints, ensuring endpoint adjacency for seamless extension. The Sierpiński curve, introduced by Wacław Sierpiński in 1915, fills an using an pattern that iteratively subdivides the triangle into four smaller congruent triangles. Starting with the first-order as a single zigzag connecting the three vertices; the second-order applies the pattern within each half, forming multiple arrowheads pointing alternately inward; and the third-order refines further, with the curve's self-similar arrow motifs densely covering the triangle in the limit while avoiding self-intersections except at endpoints. This construction suits triangular domains, with each iteration doubling the path length and enhancing fill density through . The extends the as a closed variant suitable for higher dimensions, particularly , by concatenating four copies of the Hilbert pattern in a that visits all eight octants of a without leaving gaps. In , it forms a closed from the open Hilbert path; in , iterative stages mirror Hilbert's U-patterns across layers, with first-order outlining a cubic frame, second-order filling 64 subcubes, and third-order achieving 512 subcubes for volumetric filling.
CurveLocality PreservationEase of Computation
PeanoModerate; ternary mapping allows distant points to connect across the square, leading to some jumps.High; analytical digit assignment enables direct coordinate computation without .
HilbertExcellent; recursive quadrant connections minimize distance between successive curve points, preserving neighborhoods.Moderate; requires recursive rules and rotations, implementable via iterative algorithms but more complex than linear mappings.
SierpińskiGood; arrowhead iterations maintain local within triangular subdivisions, similar to Hilbert but domain-specific.Moderate; recursive with rotational symmetries, straightforward for triangular grids but needs adaptation for squares.

Properties

Topological Properties

Space-filling curves are continuous surjective mappings from the unit interval [0,1] onto a higher-dimensional compact set, such as the unit square [0,1]^2. of the limit map arises from the uniform convergence of a of continuous polygonal approximations that refine the of the and , ensuring the supremum of the difference between successive iterates approaches zero. Surjectivity follows from the density of the images of these approximations in the target space, as every point in the is a point of the of approximations, and the uniform preserves this density to cover the entire set. The image of a space-filling curve is a Peano , characterized as a compact, connected, and locally connected . This structure aligns with the Hahn–Mazurkiewicz theorem, which states that a compact, connected, weakly locally connected is the continuous image of [0,1]. Space-filling curves are non-injective, exhibiting extensive self-overlaps where most points in the image have uncountably many preimages under the mapping. This non-injectivity prevents the curve from being a onto the unit square, as the topological of [0,1] is 1 while that of [0,1]^2 is 2, and is invariant under homeomorphisms. Topological properties of space-filling curves, including and surjectivity, remain invariant under reparameterizations of the , as such reparameterizations compose to yield equivalent mappings with the same image and behavior.

Metric and Dimensional Properties

Space-filling curves, as limits of finite polygonal approximations, possess infinite . For the , the nth-order approximation consists of $4^n line segments, each of length $2^{-n}, yielding a total length of $2^n, which diverges to as n \to \infty. Similarly, the Peano curve's approximations exhibit lengths that grow without bound, confirming the infinite length of the limiting curve. These curves are nowhere differentiable, meaning they admit no tangent line at any point. It has been proved that space-filling curves such as the Peano and Hilbert curves are nowhere differentiable (Morayne, 1987). Their reflects this irregularity, satisfying \omega(h) = O(\sqrt{|h|} \log(1/|h|)) for small h, which bounds the but prevents differentiability . The image of a space-filling curve from [0,1] to the unit square has box-counting dimension 2, matching the of the square it fills. The image of a space-filling curve has 2, matching the of the square. The graphs of its coordinate functions have 3/2. Regarding , the image occupies the full positive measure of the unit square, ensuring complete coverage. However, the preimage under the curve mapping of any point has zero, consistent with the one-dimensional parameter domain. Among space-filling curves, the excels in locality preservation, minimizing distance distortion relative to the by maintaining nearby points in the parameter domain close in the target space. This property arises from its recursive construction, which avoids large jumps unlike the more erratic traversals in the .

Characterization Theorems

Hahn–Mazurkiewicz Theorem

The Hahn–Mazurkiewicz theorem, independently established by Hans Hahn in 1914 and Stefan Mazurkiewicz in 1913, resolves a central question arising from Peano's discovery of space-filling curves by providing a precise topological characterization of those continua that can be parametrized by the unit interval. This theorem identifies the exact conditions under which a space admits a continuous surjection from [0,1], bridging the gap between one-dimensional parametrizations and higher-dimensional fillings. The theorem states that a compact X is the continuous surjective image of the unit [0,1] X is a locally connected —that is, X is compact, connected, and locally connected. Equivalently, such spaces are known as Peano continua or Peano spaces. The proof proceeds in two directions. For necessity, the continuous image of the compact interval [0,1] is compact by the closed map lemma and connected as the image of a ; moreover, local connectedness follows because [0,1] is locally path-connected, and continuous images of compact, locally path-connected metric spaces inherit local path-connectedness (hence local connectedness in the metric setting). For sufficiency, the construction relies on embedding a surjection onto a dense subset of X (possible by the local connectedness ensuring uniform local path-connectedness) and extending it continuously by connecting points with paths in the gaps, yielding a whose limit fills X. The unit square [0,1]^2 exemplifies a space satisfying the theorem's conditions: it is a compact, connected, locally connected , hence admits a space-filling curve such as Peano's or Hilbert's. In contrast, S^1 also meets the criteria and is the image of [0,1] via the standard parametrization t \mapsto (\cos 2\pi t, \sin 2\pi t), though it fills only a one-dimensional . A key negative example is the —the of countably many circles of radii $1/n shrinking to a point—which is compact, connected, and metrizable but fails local connectedness at the wedge point, precluding any continuous surjection from [0,1]. Corollaries include the absence of space-filling curves for non-locally connected continua, such as the or (in its compact closure). The theorem generalizes to the \prod_{n=1}^\infty [0,1], which is a compact, connected, locally connected and thus the continuous image of [0,1] via interleaving coordinates in ternary expansions, enabling space-filling in infinite dimensions.

Connections to Group Actions

Kleinian groups are discrete subgroups of the group of orientation-preserving isometries of hyperbolic 3-space \mathbb{H}^3, identified with \mathrm{PSL}(2, \mathbb{C}) acting via Möbius transformations on the boundary Riemann sphere \hat{\mathbb{C}}. The limit set of a Kleinian group G is the closure of the set of accumulation points of the G-orbit of any point on \hat{\mathbb{C}}, typically forming a fractal subset with Hausdorff dimension strictly between 0 and 2 for non-elementary groups. In the theory of doubly degenerate Kleinian groups, the can be the entire , and the Cannon-Thurston provides a continuous surjective from (parameterizing the boundary of the hyperbolic surface) to the , yielding a sphere-filling . These constructions, developed by James W. Cannon and William P. Thurston in the , demonstrate how group actions can produce spaces satisfying the conditions of the . For instance, Schottky groups—a class of freely generated Kleinian groups defined by pairing circles on the sphere—can generate limit sets that, in degenerate cases, lead to such surjective maps whose images fill the sphere. Post-2000 developments draw analogies via Sullivan's dictionary between Kleinian group actions and those of rational maps on the , relating to the dynamics on invariant sets like sets, though direct constructions of space-filling curves in this holomorphic setting remain limited to specific cases. The Hahn–Mazurkiewicz theorem provides the topological foundation, guaranteeing the existence of continuous surjective maps from the to connected locally compact metric spaces like these limit sets when they satisfy the conditions.

Applications

Integration Techniques

Space-filling curves provide a parametrization that allows the reduction of multidimensional s over regions like the unit square [0,1]^2 to one-dimensional integrals along the curve. For a continuous f: [0,1] \to [0,1]^2, the classical change-of-variables formula would suggest \int_{[0,1]^2} g(x,y) \, dx \, dy = \int_0^1 g(f(t)) \|f'(t)\| \, dt for a suitable integrand g, but space-filling curves are nowhere differentiable, so f' does not exist . In such cases, the integral can be approached using the Lebesgue-Stieltjes measure induced by f or through limits of linear approximations to the curve. For measure-preserving space-filling curves, such as the Hilbert or Peano curves, the pushforward of the one-dimensional Lebesgue measure on [0,1] under f coincides with the two-dimensional Lebesgue measure on [0,1]^2. This property ensures that, for Lebesgue integrable functions g, the multidimensional integral simplifies directly to \int_{[0,1]^2} g(x,y) \, dx \, dy = \int_0^1 g(f(t)) \, dt. The equality holds under suitable conditions, such as when g is Riemann integrable, by taking limits of Riemann sums along polygonal approximations of the curve that converge to the full space-filling path. In numerical quadrature, approximations of space-filling curves introduce discretization errors, where the error bound depends on the curve's Hölder continuity and the refinement level of the polygonal surrogate. For instance, using Hilbert curve approximations in quasi-Monte Carlo methods yields mean squared error rates of O(n^{-1-1/d}) for discontinuous integrands in d dimensions, improving over standard Monte Carlo's O(n^{-1}) rate. The locality property of the Hilbert curve—mapping nearby points in [0,1] to nearby subcubes in [0,1]^d—further reduces variance in Monte Carlo integration by preserving spatial correlations, leading to variance reductions comparable to grid-based sampling but applicable to arbitrary sample sizes n. Early 20th-century applications leveraged space-filling curve parametrizations to address multidimensional problems, including the numerical solution of partial differential equations by reducing them to one-dimensional traversals along the curve. In , measure-preserving variants were employed in the theory of multidimensional to simplify computations while maintaining measure-theoretic consistency. The nowhere differentiability of these curves precludes classical Riemann integration along the path without resorting to measure-theoretic extensions or approximations.

Computational Uses

Space-filling curves, such as the , are employed in spatial indexing structures like for databases handling multidimensional data. By interleaving the binary representations of coordinates, the Z-order curve maps or spatial points to a 1D sequence that preserves locality, ensuring that nearby points in the spatial domain remain adjacent in the linear order to facilitate efficient range queries and storage. This approach optimizes data organization in formats like HDF5, where quadtree partitioning combined with Z-ordering supports contiguous or chunked layouts for applications such as fluid flow simulations, reducing access times compared to unsorted structures. In image processing, the serves as a traversal order for pixels to enhance and rendering efficiency by exploiting spatial correlations and minimizing misses. For instance, in of medical images, pixels are reordered along a Hilbert path, which captures horizontal, vertical, and regional relationships more effectively than raster scans, leading to ratios up to 37% better than JPEG2000 on datasets. This locality-preserving ordering also aids in image generation by enabling recursive block-based encoding that reduces redundancy in textured or natural images. In rendering pipelines, Hilbert-based grid traversals accelerate neighbor-finding operations, improving utilization for multidimensional image data. Space-filling curves find extensive use in scientific computing for of multidimensional arrays, particularly in decomposition and data redistribution tasks from the 1990s through the 2020s. In (CFD), curves like the Peano-Hilbert and Morton orders partition Cartesian meshes across processors, enabling linear-time O(N) operations for load balancing and inter-mesh interpolation with minimal communication overhead, as demonstrated on systems scaling to 640 CPUs with 4.7 million cells achieving near-linear speedups. Implementations in MPI-based environments and GPU traversals leverage these curves to reorder arrays, reducing data movement costs in stencil computations; for example, Hilbert ordering yields cache miss rates below 0.1 per kilo-instruction on multicore CPUs, compared to over 100 for traditional layouts. Post-2010 developments have integrated space-filling curves into for embedding high-dimensional data, where Hilbert curves order features to preserve spatial locality during clustering and similarity searches, enhancing cache efficiency in algorithms like k-means and expectation-maximization on distributed systems. In geographic information systems (GIS), Hilbert curves support map tiling by encoding longitude-latitude pairs into 1D keys for key-value stores, enabling efficient range queries over tiled geospatial data and reducing scan times for spatial extents, though large ranges may still require cluster-wide operations. Compared to row-major ordering, space-filling curves offer superior preservation, maintaining spatial proximity in 1D mappings to minimize communication and misses in parallel applications; quantitative benchmarks show Hilbert and Morton orderings reducing TLB misses by orders of magnitude on slab-like data surfaces. Iterative constructions of these curves facilitate algorithmic implementations in software libraries. For generating Hilbert indices, a recursive traversal algorithm can be used, as follows (adapted for base case):
function generatePath(P, c, M, h, r):
    if r == 0:
        append c to P
    else:
        for i in 0 to 3:  // 2^d subcubes
            s = subcube center from i
            c_new = M * s + c
            h_new = [transformation](/page/Transformation) gamma_i(h)  // Apply [rotation](/page/Rotation)/[reflection](/page/Reflection)
            M_new = M * gamma_i  // Update orientation matrix
            generatePath(P, c_new, M_new, h_new, r-1)
This builds the curve by recursively subdividing and applying orientation-preserving transformations based on Gray-coded subcubes.

References

  1. [1]
    [2501.04705] Space-Filling Curves - arXiv
    Dec 23, 2024 · We examine space-filling curves, which are surjective continuous maps from [0,1] to some higher-dimensional space, usually the unit square [0,1]^2.
  2. [2]
    [PDF] Section 44. A Space-Filling Curve
    Apr 12, 2017 · A space-filling curve, also called a Peano curve, is a continuous mapping from the interval [0,1] to the unit square [0,1] x [0,1].
  3. [3]
    [PDF] Hilbert's Space Filling Curves and Geodesic Laminations
    The first space filling curve was introduced by Peano [6] in 1890. The drawing of this curve was shown in [7]. During the following year Hilbert [3] ...
  4. [4]
    Finding keys to the Peano curve - SpringerLink
    Jul 5, 2022 · The purpose of this paper is to give a fairly complete analysis of Peano's space filling curve. We begin with a synopsis of the history of the curve.
  5. [5]
    Pandimensional Space-Filling Curves - People | MIT CSAIL
    A space-filling curve is a parameterized, injective function which maps a unit line segment to a continuous curve in the unit square, cube, hypercube, etc.
  6. [6]
    [2008.01684] Space-filling Curves for High-performance Data Mining
    Aug 4, 2020 · They have numerous applications like search structures, computer graphics, numerical simulation, cryptographics and can be used to make various ...
  7. [7]
    [PDF] Empirical Analysis of Space–Filling Curves for Scientific Computing ...
    Abstract—Space-Filling Curves are frequently used in parallel processing applications to order and distribute inputs while pre- serving proximity.
  8. [8]
    [PDF] Topology Dependent Space Filling Curves for Sensor Networks and ...
    The narrow definition of space filling curve, in mathematical analysis, refers to a curve whose range contains the entire 2- dimensional unit square (or more ...
  9. [9]
    Space-Filling Curves - Book - SpringerLink
    The subject of space-filling curves has fascinated mathematicians for over a century and has intrigued many generations of students of mathematics.
  10. [10]
    [PDF] Sur une courbe, qui remplit toute une aire plane
    Sur une courbe, qui remplit toute une aire plane. Peano, pp. 157 - 160. Terms ... The PDF-version contains the table of contents as bookmarks, which allows easy ...
  11. [11]
    A Few of My Favorite Spaces: Space-Filling Curves
    Mar 31, 2016 · Space-filling curves challenge my intuition dimension. A space-filling curve is the image of a line, a fundamentally 1-dimensional object, that ...Missing: source | Show results with:source
  12. [12]
    The Journey to Define Dimension | Quanta Magazine
    Sep 13, 2021 · These are the first five steps of the process that will produce a space-filling curve. At each step the curve has zero area, but in the limit, ...
  13. [13]
    Cantor's Cardinal and Ordinal Infinities - jstor
    unit square has the same cardinality as the unit interval, and, as we have seen, the issue of continuity was explicitly discussed between Cantor and.
  14. [14]
    Sur une courbe, qui remplit toute une aire plane
    Peano, G. Sur une courbe, qui remplit toute une aire plane. Math. Ann. 36, 157–160 (1890). https://doi.org/10.1007/BF01199438Missing: space- filling curve
  15. [15]
    Ueber die stetige Abbildung einer Line auf ein Flächenstück
    Ueber die stetige Abbildung einer Line auf ein Flächenstück. Published: September 1891. Volume 38, pages 459–460, (1891); Cite this article. Download PDF.
  16. [16]
    A Jordan Curve of Positive Area - jstor
    A Jordan curve is a continuous plane curve without multiple points, defined by a one-to-one mapping to a line segment or circle. The paper questions if its ...
  17. [17]
    [PDF] Space-filling Curves --------~--------
    Examples of such 'space-filling' curves were later constructed by. Hilbert (in 1891), Moore (in 1900), Lebesgue (in 1904),. Sierpinski (in 1912) and Schoenberg ...Missing: Osgood Wazewski
  18. [18]
    [PDF] Dimensions of the coordinate functions of space-filling curves
    Sep 28, 2006 · [21] G. Peano, 'Sur une courbe qui remplit toute une aire plane', Math. Annln. 36 (1890) 157-160.
  19. [19]
    [PDF] Space-Filling Curves - Refubium
    2.1.3 History of Space-Filling Curves. In 1877/78 Cantor [28, 30] demonstrated that there is a bijective function be- tween any two finite-dimensional smooth ...Missing: Osgood Wazewski
  20. [20]
    [PDF] arXiv:1812.00766v1 [math.GN] 27 Nov 2018
    Nov 27, 2018 · [6] Peano, G.: Sur une courbe qui remplit toute une aire plane. Math. Annln. 36, 157–160 (1890). [7] Sagan, H.: Space-filling curves.
  21. [21]
    [PDF] arXiv:2501.04705v1 [math.HO] 23 Dec 2024
    Dec 23, 2024 · In this paper, we will look at Hilbert's simplified version of the Peano curve, and Lebesgue's curve, another space filling curve [0,1] → [0,1]2 ...
  22. [22]
    [PDF] arXiv:2110.01069v2 [math.DS] 4 Nov 2021
    Nov 4, 2021 · Space filling curves are a concept introduced by Peano in 1890 [Pea90], simplified by the Hilbert curve in 1891 [Hil91]. Since then many other ...
  23. [23]
    [PDF] Construction, Transformation and Structures of 2x2 Space-Filling ...
    Feb 21, 2025 · Keywords: Space-filling curve, 2x2 curve, Hilbert curve, Curve structure ... Hilbert curve (Hilbert, 1891). In it, self-similarity is ...
  24. [24]
    [PDF] The exact measures of the Sierpi´nski d-dimensional tetrahedron in ...
    From the original work of Sierpinski, what is known as the Sierpinski arrowhead curve is a continuous map of the line segment [0, 1] ⊂ R whose image is a ...
  25. [25]
    [PDF] Fractal Geometry: Historical and Conceptual Background
    Jul 15, 2025 · In 1915. Sierpinski [43] has devised a self-similar curve having almost all points as the branching points. This. Preprints.org (www.preprints.
  26. [26]
    [PDF] Harmonious Hilbert curves and other extradimensional space-filling ...
    Nov 1, 2012 · Figure 4 shows the next level of refinement of the two-dimensional Hilbert curve and the three-dimensional Butz-Moore curve. The order in which ...
  27. [27]
    [PDF] The Complex Dimensions of Space-Filling Curves - eScholarship
    Space-filling curves, often called 'fractals', are topologically simple. This dissertation explores their complex dimensions, which are linked to singularities ...Missing: source | Show results with:source
  28. [28]
    [PDF] Space-filling curves for Partitioning Adaptively Refined Meshes
    These curves are locality preserving: successive points on a curve are grid ... Some of the popular SFCs include Hilbert, Peano and. Sierpinski [3] ...<|control11|><|separator|>
  29. [29]
    [PDF] ON SPACE-FILLING CURVES AND THE HAHN-MAZURKIEWICZ ...
    There is an iterative construction of Lebesgue's curve similar to Peano's curve. It is given in figure 2. 4. The Hahn-Mazurkiewicz theorem. In this section we ...
  30. [30]
  31. [31]
    Length of Hilbert Curve in 3 Dimensions - Math Stack Exchange
    Nov 28, 2016 · The Hilbert Curve is a continuous space filling curve. The length of the nth iteration in two dimensions can be calculated by 2n−12n. The curve ...3D Hilbert Curve Without Double-Length Edges?Limit of arc-length of a curve - Mathematics Stack ExchangeMore results from math.stackexchange.comMissing: infinite | Show results with:infinite
  32. [32]
    [PDF] Infinite Numerical Computing Applied to Hilbert's, Peano's, and ...
    The Moore curve is an interesting variant of the Hilbert curve emanating ... three-dimensional Hilbert space-filling curves. arXiv: 1109.2323v2 (2016).
  33. [33]
    [PDF] A Jordan Curve of Positive Area
    May 16, 2013 · A Jordan curve is a continuous plane curve without multiple points. The paper shows that the exterior area of such a curve can be positive, not ...
  34. [34]
    Regularity and modulus of continuity of space-filling curves
    Apr 10, 2019 · We study critical regularity assumptions on space-filling curves that possess certain modulus of continuity.
  35. [35]
    Is the Fractal Dimension of a Space-Filling Curve in a Plane Always 2?
    Apr 29, 2013 · Since the Hausdorff dimension of the unit square is 2, then yes, by definition the dimension of a curve filling the unit square is 2. Share.What is the topological dimension of the Peano curve?Is it possible to construct a smooth curve with fractional Hausdorff ...More results from math.stackexchange.com
  36. [36]
    A new method of estimation of the box-counting dimension of ...
    Aug 7, 2025 · The method is based on a quasi-inverse to a space-filling curve which maps data to one dimension. Such a transformation retains box-counting ...
  37. [37]
    Proof that no differentiable space-filling curve exists - MathOverflow
    Mar 29, 2015 · Could someone provide a reference or a sketch of a proof that no differentiable space-filling curve exists? Or piecewise differentiable? Must ...Missing: Osgood 1903
  38. [38]
    [PDF] Analysis of the Clustering Properties of the Hilbert Space-Filling Curve
    Peano, in 1890, discovered the existence of a continuous curve which passes through every point of a closed square [21]. According to Jordan's precise notion ( ...<|control11|><|separator|>
  39. [39]
    The Beauty of Space-Filling Curves: Understanding the Hilbert Curve
    Sep 7, 2025 · Space-filling curves (SFC) are fascinating mathematical constructs with many practical applications in data science and data engineering.
  40. [40]
    On the Hahn--Mazurkiewicz theorem
    Nowadays by the Hahn-Mazurkiewicz theorem we understand the fol- lowing proposition. ProposiTion(H-M). Every locally connected metrizable continuum is a.
  41. [41]
    ON THE HAHN-MAZURKIEWICZ THEOREM IN NONMETRIC ...
    About 1914 H. Hahn [2] and S. Mazurkiewicz [4]1 topologically characterized continuous images of the real line segment as metric.
  42. [42]
    Shape injectivity of the earring space (Part II) - Wild Topology
    Aug 6, 2019 · Hahn-Mazurkiewicz Theorem: A space X is a Peano continuum if and only if it is Hausdorff and there is a continuous surjection ; First, consider ...
  43. [43]
    Are there space filling curves for the Hilbert cube? - MathOverflow
    Aug 9, 2010 · There is a surjective continuous map [0;1]→[0;1]2 ("space filling curve"). Using such a map one can easily get space filling curves for all ...Is the Hilbert space-filling curve bijective over computable numbers?Hemispherical space filling hilbert curve - MathOverflowMore results from mathoverflow.net
  44. [44]
    [PDF] Boundaries of Kleinian groups - arXiv
    Sep 8, 2016 · The boundary. ∂VR has the property that identifying the endpoints of the edge groups results in a planar set, since this space is homeomorphic ...<|control11|><|separator|>
  45. [45]
    [PDF] Sullivan's structural stability of expanding group actions
    Abstract. In his 1985 paper Sullivan sketched a proof of his structural stability theorem for group actions satisfying certain expansion-hyperbolicity ...
  46. [46]
    [PDF] A Crash Course on Kleinian Groups - Indico [Home]
    Jun 24, 2005 · G provides a famous example of a Kleinian group whose limit set is a space filling curve. Involved in the construction of G is the fact that ...
  47. [47]
    [2209.12453] Limit sets of cyclic quaternionic Kleinian groups - arXiv
    Sep 26, 2022 · In this paper, we consider the natural action of \mathrm{SL}(3, \mathbb{H}) on the quaternionic projective space \mathbb{P}_{\mathbb{H}}^2.Missing: filling | Show results with:filling
  48. [48]
    [PDF] The Sullivan dictionary and Bowen-Series maps
    These. Kleinian groups arise out of pinching finitely many disjoint, simple, closed curves. (on the surface D/Γ0) out of an explicit finite list. In ...
  49. [49]
    [PDF] Extensible grids: uniform sampling on a space-filling curve - Art Owen
    The space-filling curve reduces a multidimensional integral to a one-dimensional numerical integration problem. It is important to point out that the ...
  50. [50]
    Extensible Grids: Uniform Sampling on a Space Filling Curve
    The space filling curve reduces a multi-dimensional integral to a one-dimensional numerical integration problem. It is important to point out that the ...
  51. [51]
    [PDF] A Multi-dimensional Data Storage Using Quad-tree and Z-Ordering
    Abstract—Multi-dimensional applications use tree structure to store data and space filling curves to traverse data. Most frequently used Quad-tree and ...
  52. [52]
    [PDF] Fractal-Based Heuristic Pixel Segmentation for Lossless ... - arXiv
    To enhance the different mul- tidimensional patterns in medical image data, a Hilbert-curve [Fig. 2] by Cerveny, 2021 fractal pixel ordering is employed [19].Missing: rendering misses
  53. [53]
    [PDF] Efficient Neighbor-Finding on Space-Filling Curves - arXiv
    Oct 18, 2017 · Of course, multi-level tables require more cache memory to be stored and thus can impact the cache performance of a grid traversal. Example ...
  54. [54]
    [PDF] Applications of Space-Filling Curves to Cartesian Methods for CFD
    Locality: In the U-order, each cell visited by the curve is directly connected to two face-neighboring cells which remain face-neighbors in the one dimen-.
  55. [55]
    The Impact of Space-Filling Curves on Data Movement in Parallel ...
    Jul 15, 2023 · This work examines the hypothesis that space-filling curves, such as Hilbert and Morton ordering, can improve data locality and hence result in more efficient ...
  56. [56]
    A Map Tile Data Access Model Based on the Jump Consistent Hash ...
    Dec 6, 2022 · However, although a Hilbert curve can improve the efficiency of some spatial query operations to some extent, a Hilbert curve cannot group ...
  57. [57]
    [PDF] Sixteen space-filling curves and traversals for d-dimensional cubes ...
    Sep 12, 2018 · Figure 1: A sketch of Hilbert's space-filling curve. the traversals within the cells, one can ensure that each cell's first subcell touches the ...