Implementation of linear Regression

read the csv file

wc_data <- read.csv(“D:\t\drive-download-20190806T055124Z-001\wc-at.csv”)

view the dataset

View(wc_data)

attach(wc_data)

calculate the average/mean

mean(AT)

find the rows and columns in dataset

dim(wc_data)

find the Min/Max/Mean/Median/1st qu/3rd qu —> summary statistics of data set

summary(wc_data)

check if the columns are following normal distribution or not

qqnorm(AT)

plot(Waist,AT)

correlation coefficient value for Waist and AT

cor(Waist,AT)

Implement simple linear regression Model 1

model1 <- lm(AT ~ Waist,data=wc_data) summary(model1)

pv <- predict(model1,wc_data)

pv <- as.data.frame(pv)

pv

final <- cbind(wc_data,pv) View(final)

write.csv(final,“D:\t\drive-download-20190806T055124Z-001\wc_at_model_1P.csv”)

Model 2

model2 <- lm(AT ~ log(Waist),data=wc_data) summary(model2)

pv2 <- predict(model2,wc_data) pv2 <- as.data.frame(pv2)

final <- cbind(wc_data,pv2) write.csv(final,“D:\t\drive-download-20190806T055124Z-001\wc_at_model_2P.csv”)

Model 3

model3 <- lm(log(AT) ~ Waist,data=wc_data) summary(model3)

pv3 <-predict(model3,wc_data) pv3 <- as.data.frame(pv3)

final <- cbind(wc_data,pv3) write.csv(final,“D:\t\drive-download-20190806T055124Z-001\WC_AT\wc_at_model_3P.csv”)

model 4

model4 <- lm(AT ~ sqrt(Waist),data = wc_data) summary(model4)

pv4 <- predict(model4,wc_at) pv4 <- as.data.frame(pv4)

final <-cbind(wc_data,pv4) write.csv(final,“D:\t\drive-download-20190806T055124Z-001\WC_AT\wc_at_model_4P.csv”)

predict for new set of values with model 3

newdata=read.csv(“D:\t\drive-download-20190806T055124Z-001\WC_AT\x.csv”) pvnew <- predict(model3,newdata) pvnew <- as.data.frame(pvnew) View(pvnew)

final1 <- cbind(newdata,pvnew) write.csv(final1,“D:\t\drive-download-20190806T055124Z-001\WC_AT\new_P.csv”)