R的Procrustes Analysis的站点得分

时间:2014-03-13 13:35:27

标签: r

我正在尝试在R中执行Procrustes Analysis后确定站点得分坐标。

comp.ord<-procrustes(PA.PCA.hell.trans, PCA.hell.trans, scale=T, symmetric=T, scores="sites")
comp.ord
summary(comp.ord)

在输出中我得到:

Number of objects: 235    Number of dimensions: 34 

Procrustes sum of squares:  
0.3938225 
Procrustes root mean squared error: 
0.04093703 
Quantiles of Procrustes errors:
   Min         1Q     Median         3Q        Max 
0.01880827 0.03088994 0.03657154 0.04343953 0.11136274 

Rotation matrix: (##EXTREMELY LONG. WILL SPARE YOU OF THIS)

Translation of Averages: ##Also quite long
Scaling of target:
[1] 0.778574

这些信息非常有价值,但我正在寻找在绘制图形时绘制的坐标:##我不确定如何在问题中放置图形。

plot(comp.ord)

我想要缩放,旋转排序的x,y坐标

1 个答案:

答案 0 :(得分:1)

我猜你正在使用&#34;素食主义者&#34; :

 comp.ord$X
 comp.ord$Yrot
 plot(0,0, xlim=range(c(comp.ord$X[,1], comp.ord$Yrot[,1])), ylim=range(c(comp.ord$X[,2], comp.ord$Yrot[,2])), type="n")
 points(comp.ord$X, pch=19, col="red")
 points(comp.ord$Yrot, pch=19, col="blue")
 with(compr.ord, arrows(Yrot[,1], Yrot[,2], X[,1], X[,2]))

HTH