Shiva Rajab Ghal e۱ - MS Student, Borujerd Branch, Islamic Azad University
Shahram Nakhjavan - Member of Faculty, Borujerd Branch, Islamic Azad University
Mohammad Hassan Kooshki - Member of Faculty, Borujerd Agricultural Research Station

Hatami et al., 2011). In principal components analysis the components with higher values than 1 were selected. 4 components were selected in total that justified about 86% of the total variables variance (Sonboleh Kar et al., 2011). Ward cluster analysis was used to determine genotypes relationship. Genotypes were divided into three clusters. Cluster 1 included genotypes 17, 16, 15, 13, 12, 5, 3, and 1(ks 21670, ks 21684, ks 21189, ks 21191, control Khomein, local Azna, ks 21465). Genotypes in this cluster had the best conditions in term of 100-garin weight, germination percentage, grain length and width (Hatami et al., 2011). In order to classify genotypes, ward cluster analysis was used. Genotypes were classified in two clusters; cluster 1 included genotypes 4, 13, 9, 2, 3, 6, 7, 8, 5, 12, and 1 which had the highest value in terms of plant height, number of nodes, number of lateral branches, and grain diameter (Sonbole kar et al., 2011).

white beans, principal component analysis, cluster analysis