classDS {depthTools}

Distance to the Trimmed Mean Classification Method

Description

Implementation of the classification technique based on assigning each observation to the group that minimizes the distance of the observation to the trimmed mean of the group.

Usage

classDS(xl,yl,xt,alpha=0.2)

Arguments

xl

an nxp data matrix containing the observations (samples) from the learning set by rows and the variables (genes) by columns

yl

a vector of length n containing the class each observations in xl belongs to

xt

an mxp data matrix containing the observations (samples) from the test set by rows and the variables (genes) by columns

alpha

the proportion of observations that are trimmed out when computing the mean. 0.2 by default.

Details

This classification method proceeds by first computing the alpha trimmed mean corresponding to each group from the learning set, then computing the distance from a new observation to each trimmed mean. The new sample will then be assigned to the group that minimizes such distance. At the moment, only the Euclidean distance is implemented.

Value

pred

the vector of length m containing the predicted class of observations in matrix xt

Author(s)

Sara Lopez-Pintado [e-mail] and

Aurora Torrente [e-mail]

References

Lopez-Pintado, S. et al. (2010). Robust depth-based tools for the analysis of gene expression data. Biostatistics, 11 (2), 254-264.

See Also

classTAD

Examples

  ## simulated data
  set.seed(10)
  xl <- matrix(rnorm(100),10,10); xl[1:5,]<-xl[1:5,]+1
  yl <- c(rep(0,5),rep(1,5))
  xt <- matrix(rnorm(100),10,10)
  classDS(xl,yl,xt)

  ## real data
  data(prostate)
  prost.x<-prostate[,1:100]
  prost.y<-prostate[,101]
  set.seed(1)
  learning <- sample(50,40,replace=FALSE) 
  yl <- prost.y[learning]   
  xl <- prost.x[learning,]
  training <- c(1:nrow(prost.x))[-learning]
  yt.real <- prost.y[training]
  xt <- prost.x[training,] 
  yt.estimated <- classDS(xl,yl,xt)
  yt.real==yt.estimated