应用聚类分析和主成分分析法评价洋葱新品种

对酒泉市引进的27个洋葱新品种（系）叶片形态、叶色、生育期、产量等15个因子聚类分析,按其鳞茎颜色、产量、品质综合分为5类。在此基础上,对15个因子变量观测值进行主成分分析,前4个公因子对所考察性状变异的累计方差贡献率达76.849%,较好的反映了所有性状大部分信息。根据品种在各公因子上的得分值,对27个品种综合评价,综合得分按聚类分析结果分类排序,确定建议示范推广品种。

Evaluation of New Onion Varieties Using Cluster Analysis and Principal Component Analysis Methods

Using the methods of cluster analysis methods,15 factor of introducing 27 new onions varieties were selected,such as leaf shape,leaf color,growth period,yield etc,according to its bulb color,yield,quality comprehensive divided into five categories. On this basis,the method of principal components analysis is adopted to analyze factor variables for 15 observations,the results shows that cumulative total variance contribution rate amounted to 76.849% of previous 4 common factors,which reflected the most of the characteristic information;according to the scores of all varieties in all common factor,assessing the characteristic of27 varieties,which sorted by comprehensive score and the result of cluster analysis,this results is used to determine to popularize and apply varieties.