In statistics and computational geometry, the Tukey depth [1] is a measure of the depth of a point in a fixed set of points. The concept is named after its inventor, John Tukey. Given a set of n points
l{X}n=\{X1,...,Xn\}
Tukey's depth measures how extreme a point is with respect to a point cloud. It is used to define the bagplot, a bivariate generalization of the boxplot.
For example, for any extreme point of the convex hull there is always a (closed) halfspace that contains only that point, and hence its Tukey depth as a fraction is 1/n.
Sample Tukey's depth of point x, or Tukey's depth of x with respect to the point cloud
l{X}n
D(x;l{X}n)=
inf | |
v\inRd,\|v\|=1 |
1 | |
n |
n | |
\sum | |
i=1 |
1\{vT(Xi-x)\ge0\},
where
1\{ ⋅ \}
Population Tukey's depth of x wrt to a distribution
PX
D(x;PX)=
inf | |
v\inRd,\|v\|=1 |
P(vT(X-x)\ge0),
where X is a random variable following distribution
PX
A centerpoint c of a point set of size n is nothing else but a point of Tukey depth of at least n/(d + 1).