Levelwise Coverage and Finite-Dimensional Extension of Recursive Self-Indexing
Levelwise Coverage and Finite-Dimensional Extension of Recursive Self-Indexing Standalone mathematical preprint • Manuscript version 1.0 • 12 November 2026 Szalontay Anna Veronika Mathematical verification and editorial assistance: Cormeum (OpenAI ChatGPT). Original construction and research notes authored by Szalontay Anna Veronika. Abstract We present a recursive self-indexing positional construction, its digitwise inverse based on integer division and modulo, and its exact levelwise coverage. A single interleaved digit pattern is evaluated with two weight systems: one yields a natural-number index and the other a terminating nonnegative base-b value. Each expansion retains all earlier index-value pairs and adds one integer and one fractional place. For every integer radix b ≥ 2, the construction gives a bijection between nonnegative integers and finite nonnegative base-b numerals. A second digit-interleaving map produces a reversible, unique natural-number index for any finite-dimensional product of the enumerated coordinate sets. The three-dimensional case is illustrated by 125-ary blocks derived from three base-5 digits. Claims are stated for their exact domains, distinguishing density in the real half-line from coverage of the entire real line. Keywords: recursion; self-indexing; modulo; positional representation; dense rational subset; bijection; finite-dimensional indexing. 1. Motivation: traverse, extend, preserve The construction is an indefinitely extensible sequence of exhaustive finite levels. At each step all permitted digit patterns are traversed, one more integer place and one more fractional place are opened, and every previously obtained index-value pair persists. Choosing zero in both newly opened places reproduces the earlier pair exactly. In base ten the paired positional weights are (1, 0.1), (10, 0.01), (100, 0.001), and so on. Each subsequent level refines the value grid and expands its extent without removing any older result. 2. One digit pattern, two weighted evaluations Fix an integer base b ≥ 2 and a finite level k ≥ 1. Let a₀,…,aₖ₋₁ and c₁,…,cₖ belong to {0,…,b−1}. The interleaved vector is (a₀,c₁,a₁,c₂,…,aₖ₋₁,cₖ). The same digits are evaluated in two different weight systems: E1 | Natural-number self-index E2 | Associated terminating base-b value The two formulas are weighted SUMPRODUCT evaluations of the very same digit vector. The first uses alternating even and odd nonnegative powers; the second uses positive integer powers and negative fractional powers. 3. Recursion and finite-level completeness Start from the singleton index–value state {(0,0)}. For every existing pair (i,v) at level k, append a new integer digit a and fractional digit c: E3 | Recursive extension operator Pᵦ The choice a=c=0 preserves old pairs. The k-th level has 2k digit places, hence b^(2k) patterns. Their base-b index representations are precisely 0,…,b^(2k)−1, without repetition. E4 | Complete finite-level value grid Proof. The integer digits determine u∈{0,…,b^k−1}; the fractional digits determine v∈{0,…,b^k−1}. The resulting value is u+v/b^k=(u b^k+v)/b^k. Euclidean division of m=u b^k+v uniquely recovers u and v. Thus every listed grid value occurs exactly once at a fixed level. Level k Patterns, b=10 Value interval Spacing 0 1 0 – 1 100 0 … 9.9 0.1 2 10,000 0 … 99.99 0.01 3 1,000,000 0 … 999.999 0.001 4. Infinite continuation and the covered set The levels are nested. Their union contains exactly the nonnegative numbers with terminating base-b expansions: any such numeral has finitely many nonzero positional digits and hence appears by a sufficiently large level, while every constructed value has a finite expansion. E5 | Union of levels The spacing at level k is b^(−k). Therefore Sᵦ is dense on the nonnegative real half-line. Density is distinct from enumeration of all real numbers; the present theorem concerns precisely the set Sᵦ in E5. 5. Modulo inversion The key inverse step extracts the digit at base-b position t via integer division and remainder: E6 | Digit extraction Digits at even positions of the index contribute to the integer part, while digits at odd positions contribute to the fractional part: E7 | Inverse evaluation of a natural-number index For a fixed finite index the sum has finitely many nonzero terms. Conversely, the terminating base-b digits of the value, weighted as in E1, recover the same index. The two operations are inverse bijections. 6. Finite-dimensional coverage Let n be any positive finite dimension and choose n coordinates from Sᵦ. Each coordinate occurs at some finite level; the maximum of these finitely many level indices is a common level containing the full point. E8 | Finite Cartesian products commute with this nested union Proof. The inclusion from right to left is immediate. For the converse, take any n-tuple from the left. Each coordinate belongs to some Aₖ. Taking K as the maximum of these n indices, nesting implies that every coordinate belongs to A_K. Thus the point lies on the right. This argument applies to each finite n. 7. One natural-number index, with exact inverse For distinct points to have distinct indices, perform an additional digit interleaving of the one-dimensional self-indices. Let iᵣ denote the index of coordinate r, and let dₗ(iᵣ) be its base-b digit at position l: E9 | Merge n axis-indices into a single integer N E10 | Extract the axis-indices and recover coordinates The digit places n l+r do not overlap across coordinate axes. Uniqueness of base-b numeral representation proves injectivity. Every N has a unique partition of its digit positions into n classes and so reconstructs exactly one vector of axis-indices, proving surjectivity. Composition with the one-dimensional inverse yields a bijection between Sᵦ^n and the nonnegative integers. E17 | Two-way correspondence 8. Base 125 as a three-dimensional instance The n base-b digits at a fixed level can be considered a single base-(b^n) block: E11 | Block formulation E12 | Three base-5 digits give 125 possibilities Each first-block digit x₀,y₀,z₀ independently ranges over {0,1,2,3,4}; their block value runs through all 125 values exactly once. Further blocks repeat the same rule at higher places without discarding the earlier ones. Base b^n makes the exhaustive per-level traversal transparent in general finite dimension. 9. Worked and reversible numerical examples 9.1. Binary index 13 The binary digits of 13 are 1101. The even positions contribute 1+2 and the odd positions contribute 1/4: E13 | Correspondence in the original binary table 9.2. Decimal coordinates in three dimensions The point (1, 0.1, 10) has one-dimensional base-10 self-indices (1, 10, 100). Interleaving them yields: E14 | Three coordinates packed into one natural number The digits 1 occupy positions 0, 4 and 8 of 100010001. Sorting digit positions modulo three recovers (1,10,100), and the original one-dimensional inverse restores (1,0.1,10). 9.3. Three base-5 index digits The index triple (7,3,9) forms base-125 blocks D₀=117 and D₁=26: E15 | Base-125 example 9.4. Different coordinate bases: 3, 4 and 5 Coordinate axes may use different numeral bases. For (1/3, 3/4, 4/5), the respective terminating numeral systems give self-indices (3,12,20). A common base-10 index interleaving then yields: E16 | Mixed coordinate radices The coordinate bases and the base used to interleave the axis-indices play distinct roles; the latter can be any fixed integer radix at least two. 10. Scope and conclusion The proved enumeration concerns Sᵦ and Sᵦ^n for all integer b≥2 and finite n; it also extends to products of finitely many such coordinate sets with different bases. It does not, as stated, enumerate the entire real line or full real coordinate space. Work on negative values and on square-matrix representations of spatial points remains separate from the proof here. The central construction combines exhaustive levelwise traversal, preservation of previous stages, modulo inversion and one reversible natural-number code for any finite number of coordinates. Research sources [R1] Szalontay Anna Veronika (8 November 2026). A végtelen rekurzív önindexelés, a modulo és a súlyozott szorzatösszeg egységes tétele. Research manuscript, Gyűjtő3(4).pdf, especially pp. 1–5 and associated appendices. [R2] Szalontay Anna Veronika (2026). Recursive Self-Indexing Numeral Constructions and Related Computational Experiments. Previously cited Zenodo research note (identifier provided by the author): https://doi.org/10.5281/zenodo.23171627. This standalone manuscript sets out the precisely defined algebraic part of the original research and its finite-dimensional extension. Prepared for author review prior to any upload to Zenodo.
Authors
- Anna Veronika Szalontay
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-12
- DOI
- https://doi.org/10.5281/zenodo.23983753
- Primary Topic
- Advanced Combinatorial Mathematics
- Type
- preprint