Polynomial Regression - Predict tip after X number of vodka shots

vodka_shots_tip.r

# Polynomial Regression
# Number Of Vodka Shots Vs Tip
dataset = read.csv('./shots_tip.csv')


# Let's fit the data to a Linear Regression Model
linear_regressor = lm(formula = Tip ~ NumberOfShots,
                      data = dataset)


# Let's fit the data to a Polynomial Regression
dataset$NumberOfShots2 = dataset$NumberOfShots^2
dataset$NumberOfShots3 = dataset$NumberOfShots^3
dataset$NumberOfShots4 = dataset$NumberOfShots^4
dataset$NumberOfShots5 = dataset$NumberOfShots^5
polynomial_regressor = lm(formula = Tip ~ .,
                          data = dataset)

# Visualize 
# install plot package, this will help us 
# visualize the data
install.packages('ggplot2')
library(ggplot2)

# Linear
ggplot() + 
  geom_point(aes(x = dataset$NumberOfShots, y = dataset$Tip),
             colour = 'red') + 
  geom_line(aes(x = dataset$NumberOfShots, y = predict(linear_regressor, newdata = dataset)),
            colour = 'blue') +
  ggtitle('More shots Bigger tip?') +
  xlab('Number of shots') + 
  ylab('Tip')


# Polynomial Visualization
ggplot() + 
  geom_point(aes(x = dataset$NumberOfShots, y = dataset$Tip),
             colour = 'red') + 
  geom_line(aes(x = dataset$NumberOfShots, y = predict(polynomial_regressor, newdata = dataset)),
            colour = 'blue') +
  ggtitle('[Polynomial Regression] More shots Bigger tip?') +
  xlab('Number of shots') + 
  ylab('Tip')

# Let's validate the model
# predict tip accoriding the linear model
# What will be the tip after 7.5 shots of Vodka = 37.9
tip_predict_linear = predict(linear_regressor, 
                             data.frame(NumberOfShots = 7.5))

# predict tip accoriding the polynomial model
# What will be the tip after 7.5 shots of Vodka = 20.3
tip_predict_polynomial = predict(polynomial_regressor, 
                                 data.frame(NumberOfShots = 7.5, 
                                  NumberOfShots2 = 7.5^2,
                                  NumberOfShots3 = 7.5^3,
                                  NumberOfShots4 = 7.5^4,
                                  NumberOfShots5 = 7.5^5))

# For Summary
# Our Polynomial model is by far more accurate than our linear model