# Checking if the subfolder already exists.
if (!dir.exists("data")) {
    
    # Creating a subfolder to store the data.
    dir.create(path = "./data")
}

Question 1

# Downloading the American Community Survey file.
utils::download.file(url = "https://d396qusza40orc.cloudfront.net/getdata%2Fdata%2Fss06hid.csv",
                     destfile = "./data/acs.csv")

# Downloading the code book.
utils::download.file(url = "https://d396qusza40orc.cloudfront.net/getdata%2Fdata%2FPUMSDataDict06.pdf",
                     destfile = "./data/code_book.pdf",
                     mode = "wb")

# Loading CSV file into R object.
acs <- utils::read.csv(file = "./data/acs.csv")

The American Community Survey Dataset.

# First rows of ACS and first columns.

# Let's see how it is
    # CASE: github_document
    if(!knitr::is_html_output()) {
    
        # Static table using Kable Package.
        acs %>% select(c(1:5)) %>%
            head(5) %>%
            kableExtra::kbl() %>%
            kableExtra::kable_styling()
    
    # CASE: hmtl_document
    } else {
    
        # Interactive table using DT package.
        DT::datatable(acs)
    }

Dimensions of ACS dataset:

# Number of observations and Variables
dim(acs)
## [1] 6496  188

Following the question instruction, I will only display the first 3 values.

Based on the code book provided:

  • households on greater than 10 acres, means: ACR == 3
  • sold more than $10,000 worth of agriculture products, means: AGS == 6
# According to the statement.
agricultureLogical <- acs$ACR == 3 & acs$AGS == 6

# Performing the given expression:
head(which(agricultureLogical), 3)
## [1] 125 238 262

Question 2

# Loading the JPEG package.
library(jpeg)

# Downloading the KPEG file.
utils::download.file(url = "https://d396qusza40orc.cloudfront.net/getdata%2Fjeff.jpg",
                     destfile = "./data/image_file.jpeg", mode = "wb")

# Loading the JPEG file. Following the instruction to set NATIVE as TRUE.
image <- jpeg::readJPEG(source = "./data/image_file.jpeg",
                        native = TRUE)

# Answering the quantiles 30% and 80%.
c(quantile(x = image, 0.3), quantile(x = image, 0.8))
##       30%       80% 
## -15259150 -10575416

My answer is a bit different, as warned by the question instruction.

Question 3

# Downloading the GDP data.
utils::download.file(url = "https://d396qusza40orc.cloudfront.net/getdata%2Fdata%2FGDP.csv",
                     destfile = "./data/gdp.csv")

# Downloading the Educational data.
utils::download.file(url = "https://d396qusza40orc.cloudfront.net/getdata%2Fdata%2FEDSTATS_Country.csv",
                     destfile = "./data/education.csv")

# Loading Education data.
df_education <- utils::read.csv(file = "./data/education.csv")

# First rows os Education Data.
    # CASE: github_document
    if(!knitr::is_html_output()) {
    
        # Static table using Kable Package.
        df_education %>%
            select(c(1:6)) %>%
            head() %>%
            kableExtra::kbl() %>%
            kableExtra::kable_styling()
    
    # CASE: hmtl_document
    } else {
    
        # Interactive table using DT package.
        DT::datatable(df_education)
    }
# Loading GDP data.
df_gdp <- utils::read.csv(file = "./data/gdp.csv",
                          skip = 3,                   # The first rows of this database is blank
                          header = TRUE)              # Forcing the read.csv readh the first rows as header.

# First rows of GDP data.

    # CASE: github_document
    if(!knitr::is_html_output()) {
    
        # Static table using Kable Package.
        df_gdp %>%
            head() %>%
            kableExtra::kbl() %>%
            kableExtra::kable_styling()
    
    # CASE: hmtl_document
    } else {
    
        # Interactive table using DT package.
        DT::datatable(df_gdp)
    }

We need to match those databases on the country shortcode. For this reason, we need to be sure if this country shortcode is present in both datasets. The GDP database is not cleaned.

As you can see, we do not have columns with identical names neither GDP columns with numeric.

# DATA CLEANING: Fixing the GDP to be number and removing comas.
df_gdp <- df_gdp %>%
    mutate(US.dollars. = sub(pattern = " ", replacement = "", x = US.dollars.)) %>%
    mutate(US.dollars. = gsub(pattern = ",", replacement = "", x = US.dollars.)) %>%
    mutate(US.dollars. = as.numeric(US.dollars.)) %>%
    mutate(Ranking = as.numeric(Ranking))

# DATA CLEANING: Removing unecessary columns of GDP database.
df_gdp_tidy <- df_gdp %>%
    select(-c(X.1, X.2, X.3, X.4, X.5, X.6)) %>%
    rename(country_shortcode = X, ranking = Ranking, country = Economy, gdp = US.dollars.) %>%
    mutate(gdp = as.numeric(gdp)) %>%
    na.omit() %>%
    filter(ranking %in% 1:190)

# DATA CLEANING: Using the same column name to Country.
df_education_tidy <- df_education %>%
    rename(country_shortcode = CountryCode)

After the cleaning process, I merged those datasets into one.

The 13th country in the ordered dataset in a crescent way:

# Merging Education and GDP databses.
df_gdp_edu <- merge(x = df_education_tidy, y = df_gdp_tidy) 

# Searching the 13th country in crescent order.
df_gdp_edu%>%
    select(country_shortcode, Long.Name, gdp) %>%
    arrange(gdp) %>%
    filter(order(gdp) == 13)
##   country_shortcode           Long.Name gdp
## 1               KNA St. Kitts and Nevis 767

The number of countries with GDP and Education database match.

# Counting the number of countries with match.
df_gdp_edu %>% nrow()
## [1] 189

Question 4

Investigating the Income.Group variable. I just want to see what is the “levels”.

# Classes of Income Group
unique(df_gdp_edu$Income.Group)
## [1] "High income: nonOECD" "Low income"           "Lower middle income" 
## [4] "Upper middle income"  "High income: OECD"

Now, I could calculate the average of each Income.Group level. The GDP average of High income: OECD and High income: nonOECD from Income.Group could be found below:

# Calculating the average of each Income Group Level.
df_gdp_edu %>% group_by(Income.Group) %>% 
    summarise(AVG = mean(as.numeric(ranking)))
## # A tibble: 5 × 2
##   Income.Group           AVG
##   <chr>                <dbl>
## 1 High income: nonOECD  91.9
## 2 High income: OECD     33.0
## 3 Low income           134. 
## 4 Lower middle income  108. 
## 5 Upper middle income   92.1

Question 5

Let’s see the quantiles boundaries.

# Creating the groups limits.
quantile(x = as.numeric(df_gdp_edu$ranking), c(0.0, 0.2, 0.4, 0.6, 0.8, 1.0))
##    0%   20%   40%   60%   80%  100% 
##   1.0  38.6  76.2 113.8 152.4 190.0

Following the question instructions, I need to create a CUT. To do so, I will create a new variable, ranking_group, to store these grouping classifications.

# Adding a new column to the merged dataframe.
df_gdp_edu$ranking_group <- cut(x = df_gdp_edu$ranking, breaks = quantile(x = as.numeric(df_gdp_edu$ranking), c(0.0, 0.2, 0.4, 0.6, 0.8, 1.0)))

# Printing the results of this new column.
df_gdp_edu %>% select(country_shortcode, Long.Name,gdp,ranking_group) %>% head()
##   country_shortcode                    Long.Name    gdp ranking_group
## 1               ABW                        Aruba   2584     (152,190]
## 2               AFG Islamic State of Afghanistan  20497    (76.2,114]
## 3               AGO  People's Republic of Angola 114147   (38.6,76.2]
## 4               ALB          Republic of Albania  12648     (114,152]
## 5               ARE         United Arab Emirates 348595      (1,38.6]
## 6               ARG           Argentine Republic 475502      (1,38.6]

I have already inserted this new variable to classify each country based on those quantiles. I need to count the number of countries in Lower middle income in Income.Group.

df_gdp_edu %>%
    group_by(ranking_group) %>%
    filter(Income.Group == "Lower middle income") %>%
    summarise(n = n())
## # A tibble: 5 × 2
##   ranking_group     n
##   <fct>         <int>
## 1 (1,38.6]          5
## 2 (38.6,76.2]      13
## 3 (76.2,114]       11
## 4 (114,152]         9
## 5 (152,190]        16