Research Article

A STUDY ON EFFICIENCY OF RECURSIVE ALGORITHM FOR ESTIMATING RELATIONSHIP COEFFICIENTS

Maciej Gierdziewicz, Joanna Kania-Gierdziewicz

Agricultural University of Krakow

Abstract. The aim of the work was to design a computer program which would be able to calculate relationship coefficients of a population or a subpopulation in the shortest possible time, applying directly the formula used in tabular method without creating any – permanent or temporary – relationship matrices. Data were pedigrees of 25 036 Polish Black and White bulls born in the years 1960–2000. The number of all animal ID’s found in the pedigrees was 63 264. From these data the input data set for the computer program has been created. The data set contained pedigrees of the form „animal-sire-dam“, sorted in chronological order. The number of all possible animal pairs was 2 001 135 216. The algorithm was tested on four different computers, three of them having multiprocessor architecture. The method used to calculate inbreeding and relationship coefficients involved recursive relationship function calls. The recursive method let not only cut the execution time of the program 10.7 to 18 times, but let prepare the programs for parallelization which was necessary to shorten computing time and to enable more complicated pedigree analysis and calculation of wide variety of statistics.

Keywords: parallel computing, recursive algorithm, relationship coefficient

 

This Article

Accepted: 1 Jan 2007

Published online: 8 May 2017

Accesses: 670

How to cite

Gierdziewicz, M., Kania-Gierdziewicz, J., (2007). A STUDY ON EFFICIENCY OF RECURSIVE ALGORITHM FOR ESTIMATING RELATIONSHIP COEFFICIENTS. Acta Sci. Pol. Zootechnica, 6(4), 29–36.