| eval.greedy {delt} | R Documentation |
Returns a greedy histogram. A greedy histogram is grown stagewise by minimizing an empirical risk functional.
eval.greedy(dendat, leaf, method = "loglik", minobs = NULL, bound = 0, suppo = NULL)
dendat |
n*d data matrix |
leaf |
the (maximal) number of rectangles in the partition of the histogram |
method |
"loglik" or "projec"; the empirical risk is either the log-likelihood or the L2 empirical risk |
minobs |
non-negative integer; splitting of a bin will be continued if the bin containes "minobs" or more observations |
bound |
internal |
suppo |
2*d vector of real numbers; the rectangle to be splitted; the rectangle has to contain the data |
An evaluation tree
Jussi Klemela
dendat<-sim.data(n=400,seed=5,type="mulmodII") eva<-eval.greedy(dendat,leaf=15) dp<-draw.pcf(eva,pnum=c(60,60)) persp(dp$x,dp$y,dp$z,theta=-20,phi=30)