depth.curve.Tukey {curveDepth}R Documentation

Calculate Tukey curve depth using given points


Calculates Tukey curve depth of each curve in objects w.r.t. the sample of curves in data. Calculation of partial depth of each single point can be either exact or approximate. If exact, modified method of Dyckerhoff and Mozharovskyi (2016) is used; if approximate, approximation is performed by projections on directions - points uniformly distributed on the unit hypersphere.


depth.curve.Tukey(objects, data, nDirs = 100L, subs = TRUE, fracInt = 0.5,
  fracEst = 0.5, subsamples = NULL, exactEst = TRUE, minMassObj = 0,
  minMassDat = 0)



A list where each element is a multivariate curve being a list containing a matrix coords (values, d columns).


A list where each element is a multivariate curve being a list containing a matrix coords (values, d columns). The depths are computed w.r.t. this data set.


Number of directions used to inspect the space, drawn from the uniform distribution on the sphere.


Whether to split each object into two disjunctive subsets (one for integrating and one for estimation) when computing the depth.


Portion of an object used for integrating.


Portion of an object used for estimation, maximum: 1 - fracInt.


A list indicating subsamples of points for each curve in objects. Each elemnt of the list corresponds to a single curve and should be given as a vector of the length equal to the number of points on it, with entries indicating:

  • 0 do not take the point into account at all,

  • 1 use point as a reference (i.e. for integrating) and thus calculate its depth,

  • 2 utilize point in depth calculation (i.e. for estimation).


Is calculation of depth for each reference point of the curve exact (TRUE, by default) or approximate (FALSE).


Minimal portion of the objects distribution in the halfspace to be considered when calculating depth.


minimal portion of the data distribution in the halfspace to be considered when calculating depth.


A vector of doubles having the same length as objects, whose each entry is the depth of each element of objects w.r.t. data.


Lafaye De Micheaux, P., Mozharovskyi, P. and Vimond, M. (2018). Depth for curve data and applications.

Dyckerhoff, R. and Mozharovskyi P. (2016). Exact computation of the halfspace depth. Computational Statistics and Data Analysis, 98, 19-30.


# Load digits and transform them to curves
n <- 10 # cardinality of each class
m <- 50 # number of points to sample
cst <- 1/10 # a threshold constant
alp <- 1/8 # a threshold constant
curves0 <- images2curves(mnistShort017$`0`[, , 1:n])
curves1 <- images2curves(mnistShort017$`1`[, , 1:n])
curves0Smpl <- sample.curves(curves0, 2 * m)
curves1Smpl <- sample.curves(curves1, 2 * m)
# Calculate depths
depthSpace = matrix(NA, nrow = n * 2, ncol = 2)
depthSpace[, 1] = depth.curve.Tukey(
  c(curves0Smpl, curves1Smpl), curves0Smpl,
  exactEst = TRUE, minMassObj = cst/m^alp)
depthSpace[, 2] = depth.curve.Tukey(
  c(curves0Smpl, curves1Smpl), curves1Smpl,
  exactEst = TRUE, minMassObj = cst/m^alp)
# Draw the DD-plot
plot(NULL, xlim = c(0, 1), ylim = c(0, 1),
     xlab = paste("Depth w.r.t. '0'"),
     ylab = paste("Depth w.r.t. '1'"),
     main = paste("DD-plot for '0' vs '1'"))
# Draw the separating rule
dat1 <- data.frame(cbind(
  depthSpace, c(rep(0, n), rep(1, n))))
ddalpha1 <- ddalpha.train(X3 ~ X1 + X2, data = dat1,
                          depth = "ddplot",
                          separator = "alpha")
ddnormal <- ddalpha1$classifiers[[1]]$hyperplane[2:3]
pts <- matrix(c(0, 0, 1, ddnormal[1] / -ddnormal[2]),
              nrow = 2, byrow = TRUE)
lines(pts, lwd = 2)
# Draw the points
points(depthSpace[1:n, ],
       col = "red", lwd = 2, pch = 1)
points(depthSpace[(n + 1):(2 * n), ],
       col = "blue", lwd = 2, pch = 3)

[Package curveDepth version Index]