Abstract: Exploring the differences of principal component analysis methods between two algorithms of correlation matrix and covariance matrix.In order to analyze the differences in chemical composition and cluster discussion of Dendrobium,the chemical components detected by chromatographic techniques were calculated by two principal component analysis methods.Compare the differences between the two methods and choose the appropriate principal component analysis method.The results showed that the content of 11 chemical composition in 16 Dendrobium were significantly different.The principal components extracted by the correlation matrix and the covariance matrix are different,and the clustering analysis shows that the two clustering results are also different.So,the natural product components variance analysis of proximal plants or chemical cluster,even in the research of the authenticity of medicinal material,the establishment of efficient and accurate detection means was as important as the selection of appropriate data analysis methods.

Key words: principal component analysis, chemical composition of dendrobium, chromatography