#############
R-code notes compilation for  Week 2 January 25th, 27th
Compiled by: Deepta Basak. B.Math(hons.) 1st year
#############

install.packages("tidyverse")
library("tidyverse")

mpg
#fuel economy data from 1999-2008 of popular model of cars

ggplot(data = mpg) + 
  geom_point(mapping = aes(x = displ, y = hwy))
#scatter plot -> one variable against another (numeric)
#displ -> engine displacement in litres
#hwy -> highway miles per gallon
#the above plot shows negative relationship between Engine Size and Fuel Efficiency

df = mpg[c("displ", "hwy", "class")]
#creating a dataset df containing only the variables displ, hwy and class from the dataset mpg
head(df)
#a tibble: 6x3

ggplot(data = mpg) + 
  geom_point(mapping = aes(x = displ, y = hwy, colour = class))
#added a 3rd variable called class to a 2-d plot by mapping it to an aesthetic
#aes() -> associates the names of the aesthetics with the names of the variables
ggplot(data = mpg) + 
  geom_point(mapping = aes(x = displ, y = hwy), colour = "blue")
#setting color = "blue" colors all of the points blue
ggplot(data = mpg) + 
  geom_point(mapping = aes(x = displ, y = hwy, alpha = class))
#the alpha aesthetic scales the variable class by opacity of the points
ggplot(data = mpg) + 
  geom_point(mapping = aes(x = displ, y = hwy, shape = class))
#here scaling is done by assigning different shapes to the distinct observations of the variable class
ggplot(data=mpg) + 
  geom_point(mapping=aes(x=displ, y=hwy, colour=class))+
  scale_colour_viridis_d()
#using color palette from viridis package (color blind colors)
ggplot(data=mpg) + 
  geom_point(mapping=aes(x=displ, y=hwy, colour=class))+
  scale_colour_viridis_d(option = "plasma")
#using the plasma option for viridis colors

ggplot(data=mpg) + 
  geom_point(mapping=aes(x=displ, y=hwy))+ 
  facet_wrap(~ class, nrow=2)
#facet_wrap splits the plot into subplots that each display one subset of the data, with respect to the variable class
#nrow=2 specifies to arrange all of the subplots in 3 rows
ggplot(data=mpg) + 
  geom_point(mapping=aes(x=displ, y=hwy, colour=class))+
  scale_colour_viridis_d(option = "plasma")+
  facet_wrap(~class, nrow=2)
#facets and viridis
ggplot(data=mpg) + 
  geom_point(mapping=aes(x=displ, y=hwy, colour=class))+
  scale_colour_viridis_d(option = "plasma")+
  facet_grid(drv~cyl)
#facet_grid splits the plot by a combination of two variables into subplots that each display one subset of the data

ggplot(data = mpg) + 
  geom_bar(mapping = aes(x = class))
#geom_bar creates a bar chart with the x-axis displaying the variable class and the y-axis displaying the count of each observation of class
#stat_count is the default statistical transformation of geom_bar
ggplot(data = mpg) + 
  stat_count(mapping = aes(x = class))
#replacing geom_bar with stat_count produces the same plot

table = as.data.frame(table(mpg$class))
#creates a frequency table
ggplot(data = table) +
  geom_bar(mapping = aes(x = Var1, y=Freq),
           stat="identity")
#the y-axis is mapped to the Freq variable
#for manually assigning variables to the y-axis, we have to use stat="identity"
ggplot(data = mpg) + 
  stat_summary(
    mapping = aes(x = class, y = hwy),
    fun.min = min,
    fun.max = max,
    fun = median
  )
#specifying the summary function using stat_summary

ggplot(data = mpg) +
  geom_bar(mapping = aes(x = class, color = class)) 
#color=class just colors the border of each bar with respect to the class
ggplot(data = mpg) +
  geom_bar(mapping = aes(x = class, fill = class)) + 
  scale_fill_viridis_d()
#fill=class colors the entire bar; the viridis (color blind colors) color scheme has been used
ggplot(data = mpg) +
  geom_bar(mapping = aes(x = class, fill = class)) + 
  scale_fill_viridis_d() + coord_flip()
#coord_flip() flips the x and y axes
ggplot(data = mpg) +
  geom_bar(mapping = aes(x = class, fill = class)) + 
  scale_fill_viridis_d() + coord_polar()
#coord_polar() produces a stacked bar chart with polar coordinates, also known as a coxcomb chart
ggplot(data = mpg) +
  geom_bar(mapping = aes(x = class, fill = trans)) + 
  scale_fill_viridis_d()
#fill=trans creates a stacked bar chart with each bar divided into sub-bars according to the variable trans
ggplot(data = mpg) +
  geom_bar(mapping = aes(x = class, fill = trans), position="dodge") + 
  scale_fill_viridis_d()
#position="dodge" places the overlapping objects beside one another, making it easier to compare the values
#position -> stack, fill, identity,...

ggplot(data = mpg, mapping = aes(x = class, y = hwy)) + 
  geom_boxplot()
#geom_boxplot creates a vertical boxplot
#boxplot is used to summarize data using the five-number summary
ggplot(data = mpg, mapping = aes(x = class, y = hwy)) + 
  geom_boxplot() +
  coord_flip()
#using coord_flip(), we get a horizontal boxplot
ggplot(data = filter(mpg, class == "subcompact"),
       mapping = aes(x = class, y = hwy)) + 
  geom_boxplot()+ 
  coord_flip()
#creates a boxplot specifically for the class subcompact

ggplot(data = mpg,  mapping = aes(x = cty)) + 
  geom_histogram()
#geom_histogram() produces a histogram
#first specifies a sequence of points called breaks
#then, it counts the number of observations between the breaks, called bins
#a bar is placed in each bin with the height proportional to the frequency or the number of observations
ggplot(data = mpg,  mapping = aes(x = cty,fill=class)) + 
  geom_histogram() +
  scale_fill_viridis_d() 
#creates a stacked histogram with viridis colors
ggplot(data = mpg,  mapping = aes(x = cty,fill=class)) + 
  geom_histogram(binwidth = 5) +
  scale_fill_viridis_d()
#manually specifying the binwidth, which is the width between two breaks