Phylogenetic morphometrics (II): algorithms for landmark optimization

Authors

  • Pablo A. Goloboff,

    Corresponding author
    1. Consejo Nacional de Investigaciones Científicas y Técnicas, Miguel Lillo 251, 4000 S.M. de Tucumán, Argentina
    2. Instituto Miguel Lillo, Facultad de Ciencias Naturales, Miguel Lillo 205, 4000 S.M. de Tucumán, Argentina
    Search for more papers by this author
  • Santiago A. Catalano

    1. Consejo Nacional de Investigaciones Científicas y Técnicas, Miguel Lillo 251, 4000 S.M. de Tucumán, Argentina
    2. Fundación Miguel Lillo, Miguel Lillo 251, 4000 S.M. de Tucumán, Argentina
    Search for more papers by this author

  • Even better costs are possible; for the points U = (9.71376, 0.44698), V = (9.71376, 1.44698), W = (10.00, 1.00), the total tree cost is 21.75505—this is the best cost we were able to find for this case (rounded to 5 decimals).

Corresponding author:
E-mail address:pablogolo@csnat.unt.edu.ar

Abstract

This paper describes algorithms for optimizing two- or three-dimensional landmark data onto trees directly. The method is based on a first approximation using grids, and subsequent iterative refinement of the initial point estimates. Details of the implementation are discussed, as well as an empirical example.

© The Willi Hennig Society 2010.

Ancillary