clues (version 0.6.2.2)

Curve: Curve Data Set

Description

A toy example used to illustrate curve clustering.

Usage

data(Curve)

Format

A list contains a 300 by 10 data matrix (curve) and a 300 by 1 cluster membership vector (curve.mem). There are 3 clusters, each containing 100 data points, respectively, in a 10-dimensional space.

Details

cluster one is generated from the model: $$y_{ik} = sin(2 * PI * x_k) + e_{ik}, x_k \sim N(0, 1), e_{ik} \sim N(0, 0.1), i = 1, \ldots, 100, k = 1, \ldots, 10.$$

cluster two is generated from the model: $$y_{ik} = cos(2 * PI * x_k) + e_{ik}, x_k \sim N(0, 1), e_{ik} \sim N(0, 0.1), i = 1, \ldots, 100, k = 1, \ldots, 10.$$

cluster three is generated from the model: $$y_{ik} = e_{ik}, e_{ik} \sim N(0, 1), i = 1,\ldots, 100, k = 1, \ldots, 10.$$

Examples

# NOT RUN {
    data(Curve)
 
    # data matrix
    curve <- Curve$curve
 
    # 'true' cluster membership
    curve.mem <- Curve$curve.mem
    # 'true' number of clusters
    nClust <- length(unique(curve.mem))
 
    # plot average trajectories
    plotAvgCurves(curve, curve.mem)

# }