# Confidence ellipse in PCA

Is there any good explanation about the confidence ellipse used in PCA analysis which briefly describes how to interpret the result? For example, I am looking for the following cases:

• Some points are not included in the ellipse. What does that mean? How about few and large number of observations (individuals) are not included in the ellipse?

• What does the angle of the ellipse reveal? For example, what is the difference between a horizontal ellipse (stretched over PC1) versus angular ellipse versus a vertical ellipse (stretched over PC2).