Exceptional projections of sets exhibiting almost dimension conservation

Ryan E. G. Bushling
DOI: https://doi.org/10.54330/afm.143289
2024-07-01
Abstract:We establish a packing dimension estimate on the exceptional sets of orthogonal projections of sets satisfying an almost dimension conservation law. In particular, the main result applies to homogeneous sets and to certain graph-directed sets. Connections are drawn to results of M. Rams and T. Orponen.
Classical Analysis and ODEs,Metric Geometry
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the packing dimension estimation of the exceptional sets of orthogonal projections. Specifically, the paper focuses on how to estimate the packing dimension of the exceptional sets when sets satisfying the almost - dimension - conservation law are orthogonally projected. Here, the "exceptional sets" refer to those sets whose dimensions significantly decrease after being projected in specific directions. ### Background and Problem Description In geometric measure theory, a central problem is to study the dimension changes of sets under different projections. For example, Marstrand's projection theorem gives the boundary estimation of the Hausdorff dimension of sets under orthogonal projections. However, for some special cases, such as highly regular sets or specific types of graph - derived sets, these estimations may not be precise enough or applicable. ### Main Contributions of the Paper 1. **Packing Dimension Estimation**: The paper establishes the packing dimension estimation of the exceptional sets of sets satisfying the almost - dimension - conservation law under orthogonal projections. This result is applicable not only to homogeneous sets but also to some graph - derived sets. 2. **Generalize Existing Results**: The results of the paper generalize the results of Rams and Orponen et al. under specific conditions, especially when the sets are highly regular. 3. **Introduce the Concept of "Almost - Dimension - Conservation"**: The paper introduces the concept of "almost - dimension - conservation", which is more relaxed than the traditional "dimension - conservation" but can still provide useful dimension estimations. ### Main Theorems - **Theorem 1.5**: Let \(A\subset\mathbb{R}^n\) be a bounded set, and \(\dim A = \dim_B A\). If there exists a set \(F\subseteq\text{Gr}(n,k)\) such that \(t:=\dim_P F\), and for all \(V\in\text{Gr}(n,k)\setminus F\), the projection \(\pi_V\) is almost dimension - conserving for \(A\), then: \[ \dim_P\{V\in\text{Gr}(n,k):\dim A_V\leq s\}\leq\max(k(n - k)-(k - s),t)\quad\forall 0\leq s<\dim A. \] - **Corollary 1.6**: Let \(K\subset\mathbb{R}^n\) be a homogeneous set or a graph - derived set, and satisfy one of the following two properties: (1) its transformation group is finite; (2) the action of the transformation group on \(\text{Gr}(n,k)\) has a dense orbit. Then: \[ \dim_P\{V\in\text{Gr}(n,k):\dim K_V\leq s\}\leq k(n - k)-(k - s)\quad\forall 0\leq s<\dim K. \] ### Methods and Techniques - **Discretization Technique**: The paper uses the discretization technique to deal with point sets on the Grassmannian, thus simplifying the complexity of the problem. - **Energy Estimation**: By defining the energy function and estimating the upper and lower bounds, the paper proves the main theorems. - **Projection Theory**: Using the properties of orthogonal projections, especially the special properties on homogeneous sets and graph - derived sets, the paper obtains the key dimension estimations. ### Conclusion By introducing the concept of "almost - dimension - conservation", the paper successfully extends the existing projection dimension estimation results, providing new tools and methods for studying set projections in high - dimensional spaces. This result is not only of great theoretical significance but also provides a new perspective for set analysis in practical applications.