Description Usage Arguments Value References Examples
This function takes a matrix X defining the same n individuals and a vector y defining also those individuals.
1 2 3 | outlierSIR(X, y, method = c("BOOT", "TTR", "MONO"), kernel = "normal",
bandwidth = NULL, H = 5, nb.replications = 1000, pourcent = 0.1,
plotBandwidthEst = F, NCORES = 1)
|
X |
A matrix nxp describinf n observations through p varaibles. |
y |
A vector of n observations corresponding to the n rows of X. |
method |
Which method to be performed. One of BOOT for bootstrap, TTR for train-test replication and MONO for mono-model testing. |
kernel |
The kernel to be used, one of "normal" or "box". Default is "normal". Can be abbreviated. See |
bandwidth |
A real indicating the width of the kernel. If NULL, automatic kernel width tuning is used. Default is "NULL". |
H |
A positive real value. The number of slices in the SIR model. Default is 5. |
nb.replications |
An interger giving the number of repliccations for BOOT and TTR methods. Default is "1000". |
pourcent |
A real between 0 and 1. The proportion of observations to keep in the test dataset. Default is "0.1". |
plotBandwidthEst |
An integer. Wether or not to plot the EDR direction with the kernel estimation. Default is "FALSE". |
NCORES |
An integer. Fiwing the number of cores to be used in the sampling computations. Only BOOT method has been parallelized due to its computation time. |
A list containing indices of outliers in the case of the 3 methods and also the extreme observations in the case of BOOT.
1 2 3 4 5 6 7 8 9 | data(ozone)
# MONO method
output <- outlierSIR(ozone[,-c(1,12,13)],ozone[,1],method="MONO")
# TTR method
# output <- outlierSIR(ozone[,-c(1,12,13)],ozone[,1],method="TTR",
# plotBandwidthEst = FALSE,kernel="normal",nb.replications = 1000)
# BOOT method, using 5 cores of computation
# output <- outlierSIR(ozone[,-c(1,12,13)],ozone[,1],method="BOOT",
# plotBandwidthEst = TRUE,kernel="normal",nb.replications = 5000,NCORES=5)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.