top of page

Working With Prostate Cancer Dataset Using R | R Programming Assignment Help | Basic Practice Set


Use the prostate cancer dataset for the following exercises

  1. Read the dataset

  2. Examine the structure of the dataset.

  3. Remove the first variable(id) from the data set

  4. Get the number of Benign (B) cases and Malignant (M) cases. Hint: ‘table’

  5. Create a normalize function

  6. Using the function created in Question 5, normalize the numeric features in the data set.

  7. Confirm that the normalization worked

  8. Create the training(1 through 65) and test datasets (66 through 100)

  9. Use the knn() function to classify test data

  10. Evaluate the model performance


Code Implementation

##install packages

install.packages("psych")
install.packages('class')

library(class)

library(psych)

#1. Read the dataset

data <- read.csv("C:/Users/navee/OneDrive/Desktop/Oct 2022/Deadline 14 Oct +1 (202) 902-3768 R Programming/Prostate_Cancer.csv", 
                 stringsAsFactors=TRUE,sep = ",")

#2. Examine the structure of the dataset.

str(data)

#3. Remove the first variable(id) from the data set

data <- data[,-1]
head(data)

#4. Get the number of Benign (B) cases and Malignant (M) cases. Hint: 'table'

table(data["diagnosis_result"])

#5. Create a normalize function

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

#6. Using the function created in Question 5, normalize the numeric features in the data set.

data.n <- as.data.frame(lapply(data[,2:9], normalize))

#7. Confirm that the normalization worked

head(data.n)

#8. Create the training(1 through 65) and test datasets (66 through 100)

train.data <- data.n[1:65,]
test.data <- data.n[66:100,]
head(train.data)
head(test.data)

train.label <- data[1:65,1]
test.label <- data[66:100,1]
head(train.label)
head(test.label)

#9. Use the knn() function to classify test data

knn.res <- knn(train=train.data, test=test.data, cl=train.label, k=8)

#10. Evaluate the model performance

ACC.res <- 100 * sum(test.label == knn.res)/NROW(test.label)
ACC.res

Recent Posts

See All

Comments


bottom of page