# Multiple Linear Regression

rm(list = ls())

# Importing the dataset
dataset <- read.csv(file.path(getwd(),'Data/50_Startups.csv'))


# Encoding categorical data
dataset$State # to know which levels do we have in the dataset
#dataset$State = factor(dataset$State,
#                       levels = c('New York', 'California', 'Florida'),
#                       labels = c(1, 2, 3))

# i wrote the below code, i think it is more general
dataset$State <- factor(dataset$State,
                       levels = dataset$State,
                       labels = 1:length(dataset$State))


# Splitting the dataset into the Training set and Test set

if (!require("caTools"))
  install.packages("caTools") # to install the caTools package if you don't have it
set.seed(123)
split <- sample.split(dataset$Profit, SplitRatio = 0.8)
training_set <- subset(dataset, split == TRUE)
test_set <- subset(dataset, split == FALSE)

# no need for Feature Scaling with regressio

# Fitting Multiple Linear Regression to the Training set
regressor <- lm(formula = Profit ~ .,
               data = training_set)

# Predicting the Test set results
y_pred <- predict(regressor, newdata = test_set)

# Building the optimal model
# this requires to select the best variables
# the code provided throught the video is stupid because the approach is manual for selecting variables
# therefore, i found a good way to select the best variables


regressor_1 <- lm(formula = Profit ~ Administration, data = dataset) # only one independent variable
regressor_full <- lm(formula = Profit ~ ., data = dataset) # all the independent variables


# The forward variable selection for regression
step(regressor_1, scope = list(lower = regressor_1, upper = regressor_full),
     direction="forward")


# The backward variable selection for regression
step(regressor_full, data = dataset, direction = "backward")


# The stepwise (also known as Bidirectional) variable selection for regression
step(regressor_1, scope = list(upper = regressor_full), data = dataset, direction="both")


##Ploting##

# plot of the fly
plot(dataset, pch = 16, col = "blue", main = "Title")

# Finding the correlation among variables
if (!require("corrplot"))
  install.packages("corrplot") # to install the corrplot package if you don't have it
data_cor <- cor(data.frame( c(dataset[1:3], dataset[5]))) # i had to remove the state column
corrplot(data_cor, method = "number")

# Ploting the results of regression
plot(regressor, pch = 16, col = "blue", main = "Title")

# The above code will give many plots, therefore, if we want a specific plot we can use:
plot(regressor, pch = 16, which = 1, col = "blue", main = "Title")