#step 1 Import Data

setwd("~/NYU/classes/2. R/Assignments/Lesson 5")
library(readr)
le <- read_csv("life_expectancy.csv")
## Rows: 264 Columns: 60
## -- Column specification --------------------------------------------------------
## Delimiter: ","
## chr  (2): Country Name, Country Code
## dbl (57): 1960, 1961, 1962, 1963, 1964, 1965, 1966, 1967, 1968, 1969, 1970, ...
## lgl  (1): 2017
## 
## i Use `spec()` to retrieve the full column specification for this data.
## i Specify the column types or set `show_col_types = FALSE` to quiet this message.
View(le)
gdp <- read_csv("gdp.csv")
## Rows: 264 Columns: 60
## -- Column specification --------------------------------------------------------
## Delimiter: ","
## chr  (2): Country Name, Country Code
## dbl (58): 1960, 1961, 1962, 1963, 1964, 1965, 1966, 1967, 1968, 1969, 1970, ...
## 
## i Use `spec()` to retrieve the full column specification for this data.
## i Specify the column types or set `show_col_types = FALSE` to quiet this message.
View(gdp)

#Step #2 - merging the data

country= gdp$`Country Name`
gdp2016= gdp$`2016`/1000000000000
le2016=le$`2016`

#step #3 create a new data frame

gdp_le <- data.frame(country,gdp2016,le2016)
View(gdp_le)

#step # 4 create a scatter plot

plot(gdp_le$gdp2016, gdp_le$le2016)

#step 5 - refine the plot

#same plot function, 
#add x axis title, 
#add y axis title (ylab=y lable), 
#main = chart title, 
#remove the box (frame.plot), 
#change color (col=), 
#lastly set limits  0 - 100. (the last figure looks to be out of the parameters)

plot(gdp_le$gdp2016, gdp_le$le2016, xlab="GDP (in trillions of USD)", ylab="Life expectancy", main = "GDP and life expectancy in 2016 by country", frame.plot = FALSE, col="#4cbea3", xlim=c(min=0, max=100))

#Task #2 - Create a histogram of GDP

hist(gdp_le$gdp2016, main = "GDP in 2016 by country", xlab = "GDP (in trillions of USD)", col="red", breaks=10)

#Task #3a: Create a box ans whisker xplot of Life Expectancy

boxplot(gdp_le$le2016, main = "Life expectancy in 2016 by country", xlab = "Life expectancy", ylab = "Age", col="green")