A numerical exploration of the relationship between deterministic dynamics, geometry, and arithmetic properties
Understand the Model
Explore the Experiments
Review the Code and Data
The idea in a few lines
Particles numbered 1,2,3,… are released one at a time from the origin of a lattice.
Each site contains a rotor indicating a direction. A particle follows the rotor instructions until it reaches a site that has not yet been occupied. It then settles permanently at that site.
The process is fully deterministic: once the rules and the initial configuration have been fixed, the result can be reproduced exactly.
If particle n settles at site x, we write
$f(n)=x$.
Conversely, if site x is occupied by particle n, we define
$T(x)=n$.
The map $f$ tells us where a particle settles. The field $T$ tells us which particle occupied a given site.
This second representation is the central object studied on this website. It makes it possible to project arithmetic properties of integers onto the lattice, including parity, divisibility, congruence classes, decimal digits, primality, and other numerical filters.
Research questions
The occupation-rank field reveals structures that are not immediately visible in the overall shape of the aggregate.
The project investigates questions such as:
- how are congruence classes distributed throughout the aggregate?
- which effects depend on the cyclic order of the rotors?
- how is a particle number related to its path?
- which particles return to the origin after being released?
- do digit-based filters generate measurable hierarchical structures?
- do the patterns observed for prime numbers remain after parity and congruence effects have been controlled?
These questions are studied through numerical simulations on the square lattice and, in later work, on other lattices.
An experimental and reproducible approach
This website presents an independent numerical research project.
Each statement is identified according to its scientific status:
Definition
A precisely specified rule or mathematical object.
Measurement
A value computed directly from the data.
Numerical observation
A phenomenon observed in one or more simulations.
Conjecture
A statement suggested by the experiments but not proved.
Unless explicitly stated otherwise, the structures presented on this website should not be interpreted as theorems.
The conventions, parameters, algorithms, and data required to reproduce the experiments are progressively released with the results.
Research programme
1. Foundations
Precisely define the model, the placement map f, the occupation-rank field T, and the numerical protocol.
2. Congruence classes
Measure the spatial distribution of the classes n≡a(modm) and determine how these distributions evolve as N increases.
3. Particle paths and returns to the origin
Study path lengths and the observed relationship between the absence of a return to the origin and indices that are not divisible by 4.
4. Digits and hierarchical patterns
Analyse the patterns obtained by selecting sites according to the digits of their occupation ranks, in several numerical bases.
5. Primality
Search for a possible prime-specific effect after controlling for parity, small congruences, and the changing density of prime numbers.
Why publish this work online?
This website has three main purposes:
- to make the experiments understandable;
- to make them reproducible;
- to invite critical review.
Corrections, counterexamples, relevant references, independent reproductions, and alternative explanations are especially valuable.
A negative result or an identified error is also scientifically useful. Pages will therefore be corrected and versioned whenever new evidence becomes available.
Contribute to the critical review
Useful contributions include:
- identifying an ambiguous definition;
- reporting an error in a figure or calculation;
- reporting a reproducibility problem;
- suggesting a relevant scientific reference;
- proposing an alternative explanation;
- providing a counterexample to a conjecture;
- improving the simulation algorithm.
Suggested button: Submit Feedback
About
This project is developed by Jean-Pierre Coursault as an independent numerical investigation of rotor-router models.
The initial experiments focused on the placement of several million particles, arithmetic filters, and particle-path analysis. The work is now being reorganised into methodological notes and more focused thematic studies.
The observations and conjectures presented on this page may evolve in light of new simulations or critical feedback.
