R: K nearest neighbours test

prostate_cancer.R

setwd("~/R/harjottelut/prostate cancer")
library(class)
library(gmodels)
prc <- read.csv("Prostate_Cancer.csv", stringsAsFactors = FALSE)
#str(prc)
prc <- prc[-1]
#View(prc)
#table(prc$diagnosis_result)

prc$diagnosis_result <- factor(prc$diagnosis_result, levels = c("B", "M"), labels = c("Benign", "Malignant"))
#round(prop.table(table(prc$diagnosis)) * 100, digits = 1)

normalize <- function(x) {
  return ((x - min(x)) / (max(x) - min(x))) 
}

prc_n <- as.data.frame(lapply(prc[2:9], normalize))
#View(prc_n)
#summary(prc_n$radius)

prc_train <- prc_n[1:65,]
prc_test <- prc_n[66:100,]

prc_train_labels <- prc[1:65, 1]
prc_test_labels <- prc[66:100, 1]

prc_test_pred <- knn(train = prc_train, test = prc_test, cl = prc_train_labels, k=12) #decided to go with k=12

CrossTable(x = prc_test_labels, y = prc_test_pred, prop.chisq = FALSE) #shows the results. not perfect, but decent