We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cond-mat.stat-mech

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo

Condensed Matter > Statistical Mechanics

Title: Comment on "Universal and accessible entropy estimation using a compression algorithm"

Abstract: In a recent Letter [1] a framework for estimating entropy was introduced and applied to one-dimensional and two-dimensional systems. In this Comment we show that the method is not well suited for estimating entropy in bidimensional systems presenting long-range correlations.
Comments: 2 pages, 1 figure
Subjects: Statistical Mechanics (cond-mat.stat-mech); Data Analysis, Statistics and Probability (physics.data-an)
Journal reference: Physical Review Letters 129, 029801 (2022)
DOI: 10.1103/PhysRevLett.129.029801
Cite as: arXiv:2207.06868 [cond-mat.stat-mech]
  (or arXiv:2207.06868v1 [cond-mat.stat-mech] for this version)

Submission history

From: Edgardo Brigatti [view email]
[v1] Thu, 14 Jul 2022 12:45:47 GMT (457kb,D)

Link back to: arXiv, form interface, contact.