wisc: Wisconsin Breast Cancer dataset

In danno11/SMVCIR: Sliced Mean Variance Covariance Inverse Regression

Description

Wisconsin Breast Cancer dataset

Usage

data(wisc)

Format

An object of class data.frame with 683 rows and 10 columns.

Source

UCI Machine Learning Repository

Class Variable Tumor Type "tumor": "malignant" or "benign" Attributes used:

1. Clump Thickness

2. Uniformity of cell size (1-10)

3. Uniformity of Cell Shape (1-10)

4. Marginal adhesion (1-10) 5. Single Epithelial Cell Size (1-10)

6. Bare Nuclei (1 - 10)

7. Bland Chromatin (1-10)

8. Normal Nucleoli (1-10)

9. Mitoses(1-10)

10. Type of Tumor (malignant, benign)

Examples

data(wisc)
library(caret)
train<-createDataPartition(wisc$tumor, p = .8, list = F)###Create a training set using 80% of dataset
wisc.smv<-smvcir("tumor", data = wisc[train,], test = T) ###Build smvcir model on training set
bcpreds<-predict(wisc.smv, newdata = wisc[-train,], type = "prob")
head(bcpreds)  ###probability estimates


###Get Coordinates
coords<-predict(wisc.smv, newdata = wisc, coordinates_only = TRUE)
coords$tumor<-wisc$tumor


plotSVM3d<-function(x, y, kernel = "radial", ...){
 open3d()
 plot3d(x, col = as.numeric(y)+1)
 svm_mod<-svm(x = x, y = y, kernel =paste(kernel), type = "C-classification", ...)
 n=100
 nnew = 50
 newdat.list = lapply(x, function(x) seq(min(x), max(x), len=nnew))
 newdat      = expand.grid(newdat.list)
 newdat.pred = predict(svm_mod, newdata=newdat, decision.values=T)
 newdat.dv   = attr(newdat.pred, 'decision.values')
 newdat.dv   = array(newdat.dv, dim=rep(nnew, 3))
 # Fit/plot an isosurface to the decision boundary
 contour3d(newdat.dv, level=0, x=newdat.list[[1]], y=newdat.list[[2]], z=newdat.list[[3]], add=T)
 return(list(svm_mod = svm_mod))
 }


plotSVM3d(x = coords[,1:3], coords[,10], kernel = "linear") ###Visualize a support vector machine model with our coordinates
plotSVM3d(x = coords[,1:3], coords[,10], kernel = "radial")

danno11/SMVCIR documentation built on May 14, 2019, 6:06 p.m.