#### Load libraries ###
library(ISLR)
library(plotly)
library(dplyr)

#### Pull in the Wage dataset from the ISLR package ####
data("Wage")

##### EDA ####

# Prepare a Loess smoother model for the Wage Vs Age plot
wageAgeLoessModel <- loess(wage~age, data = Wage)
wageAgeLoessPredict <- predict(wageAgeLoessModel, newdata = seq(from = min(Wage$age), to = max(Wage$age), by = 1))
wageAgeLoess <- data.frame(
  age = seq(from = 18, to = 80, by = 1),
  wage = wageAgeLoessPredict
)

# Make the plot wage vs age with the loess model overlaid
wageVsAgePlot <- plot_ly(Wage) %>%
  add_trace(x = ~age, y = ~wage, type = "scatter", mode = "markers", name = "Sample",
            marker = list(color = "darkgrey"), showlegend = FALSE) %>%
  add_trace(data = wageAgeLoess, x = ~age, y = ~wageAgeLoessPredict, type = "scatter", mode = "lines", name = "Loess Smoother",
            line = list(
              color = "darkblue",
              width = 4
            ), showlegend = FALSE)

# Prepare a lm model for the Wage Vs Age plot
wageYearLmModel <- lm(wage~year, data = Wage)
wageYearPredictData <- data.frame(year = seq(from = min(Wage$year), to = max(Wage$year), by = 1))
wageYearPredict <- predict(wageYearLmModel, newdata = wageYearPredictData)
wageYearLm <- data.frame(
  year = seq(from = min(Wage$year), to = max(Wage$year), by = 1),
  wage = wageYearPredict
)

# Make the plot wage vs year with the lm model overlaid
wageVsYearPlot <- plot_ly(Wage) %>%
  add_trace(x = ~year, y = ~wage, type = "scatter", mode = "markers", name = "Sample",
            marker = list(color = "darkgrey"), showlegend = FALSE) %>%
  add_trace(data = wageYearLm, x = ~year,  y = ~wage, type = "scatter", mode = "lines", name = "Linear model",
            line = list(
              color = "darkblue",
              width = 4
            ), showlegend = FALSE)

# Make the plot wage vs education level
wageVsEducationPlot <- plot_ly(Wage) %>%
  add_trace(x = ~education, y = ~wage, color = ~education, type = "box", showlegend = FALSE)




# Join the plots
s1 <- subplot(wageVsAgePlot, wageVsYearPlot, wageVsEducationPlot)