#How to edit name of plot in sas jmp software#
If you compare PCs from two different software packages, you might notice that a PC from one package is the negative of the same PC from another package. Note that the principal components (which are based on eigenvectors of the correlation matrix) are not unique. The fourth PC is a weighted contrast between the SepalWidth and PetalLength variables (with positive coefficients) and the SepalLength and PetalWidth variables (with negative coefficients).
In a similar way, the third PC is primarily a weighted contrast between the SepalLength and PetalWidth variables, with smaller contributions from the other variables.You can interpret this weighted sum as a vector that points mostly in the direction of the SepalWidth variable but has a small component in the direction of the SepalLength variable. Is approximately PC2 ≈ 0.38*SepalLength + 0.92*SepalWidth. For the second PC, the coefficients for the PetalLength and PetalWidth variables are very small.You can interpret this as a contrast between the SepalWidth variable and an equally weighted sum of the other variables. The first PC is the linear combination PC1 = 0.52*SepalLength – 0.27*SepalWidth + 0.58*PetalLength + 0.56*PetalWidth.The linear coefficients for the PCs (sometimes called the "loadings") are shown in the columns of the Eigenvectors table. The principal components are linear combinations of the original data variables.īefore we discuss the graph, let's identify the principal components and interpret their relationship to the original variables. Ods output Eigenvectors=EV /* to create loadings plot, output this table */ run ID id /* use blank ID to avoid labeling by obs number */ Var SepalLength SepalWidth PetalLength PetalWidth /* or use _NUMERIC_ */ Out=PCOut /* only needed to demonstate corr(PC, orig vars) */ Proc princomp data=iris /* use N= option to specify number of PCs */ STD /* optional: stdize PC scores to unit variance */