Measures of Statistical Complexity: Why?

David P. Feldman
Department of Physics
University of California
Davis, California 95616, USA
J. P. Crutchfield
Physics Department
University of California
Berkeley, California 94720-7300, USA
and
Santa Fe Institute
1399 Hyde Park Rd.
Santa Fe, NM 87501, USA

ABSTRACT: We review several statistical complexity measures proposed over the last decade and a half as general indicators of structure or correlation. Recently, Lopez-Ruiz, Mancini, and Calbet [Phys. Lett. A 209 (1995) 321] introduced another measure of statistical complexity C_LMC that, like others, satisfies the "boundary conditions" of vanishing in the extreme ordered and disordered limits. We examine some properties of C_LMC and find that it is neither an intensive nor an extensive thermodynamic variable. It depends nonlinearly on system size and vanishes exponentially in the thermodynamic limit for all one-dimensional finite-range spin systems. We propose a simple alteration of C_LMC that renders it extensive. However, this remedy results in a quantity that is a trivial function of the entropy density and hence of no use as a measure of structure or memory. We conclude by suggesting that a useful "statistical complexity" must not only obey the ordered-random boundary conditions of vanishing, it must also be defined in a setting that gives a clear interpretation to what structures are quantified.


D. P. Feldman and J. P. Crutchfield, "Measures of Statistical Complexity: Why?", Physics Letters A 238 (1997) 244-252. [ps.gz]= 112kb [ps]= 331kb [pdf]= 271kb