**Data Science: Python Final Project**

---

**Created on November 7, 2020**

**Team 19 :**

 **- Chin Hsin Hsieh**
 **- Dhananjay Sonawane**
 **- Ines Cerdan**
 **- Jose Eduardo Dominguez Navarrete**

About this Project

---

Figure 1: Arabian Peninsula Region

The dataset assigned for our group is the Arabian Peninsula region. (Figure 1) In this report, we will use the data of the World Bank for conducting various exploratory data analysis. Based on these analysis results, we will get the country who can most represent the Arabian Peninsula in the end of this report.

About Arabian Peninsula

---

The Arabian Peninsula is the largest peninsula in the world, located in the northern and eastern hemispheres in the western part of the Asian continent. It is a desert environment surrounded by saltwater bodies and bounded by the Persian Gulf, the Gulf of Aden, and the Red Sea. (Figure 2)

There are the following major countries on the Arabian Peninsula. According to the size of the country, they are Saudi Arabia, Yemen, Oman, United Arab Emirates, Kuwait, Qatar, and Bahrain. Among them, Saudi Arabia is the largest, its territory occupies 75% of the peninsula.

The Arabian Peninsula is controlled by the sub-high pressure zone and the trade wind belt all year round, and it is very dry. Almost the entire peninsula is a tropical desert climate zone with a large area of no flow and lacks natural freshwater resources. Therefore, economic activity is restricted and agricultural development is not conducive.

Despite that, the Arabian Peninsula has a large number of oil reserves along the Persian Gulf. Since the successful mining in the 20th century, it has brought huge wealth to the countries on the Arabian Peninsula adjacent to the Persian Gulf.

Figure 2: The Map of the Arabian Peninsula

Importing Dataset and Installing Essential Packages

---

In this section, we are going to import Data Packages to work with the data, and import Final Project Dataset and read it with pandas packages.

#import Data Packages to work with the data

import pandas                 as pd       # data science essentials
import numpy                  as np       # arithmetic operations essentials
import matplotlib.pyplot      as plt      # data visualization essentials
import seaborn                as sns      # enhanced data visualization

#Import Final Project Dataset and read it with pd

file  = 'Final_Project_Dataset.xlsx'
final = pd.read_excel(file)

Cleaning and Subsetting Data

---

After we finished the importing and installing process, we'll start cleaning and subsetting to narrowing down our data.

#Subsetting the Data to focus only in our region

df_wheel = final[final["Cool Name"] == "Wheeljack"]
df_bumb = final[final["Cool Name"] == "Bumblebee"]
df_rat  = final[final["Cool Name"] == "Ratchet"]
df_mir  = final[final["Cool Name"] == "Mirage"]
df_side = final[final["Cool Name"] == "Sideswipe"]
df_sm   = final[final["Cool Name"] == "Skids & Mudflap"]
df_iron = final[final["Cool Name"] == "Ironhide"]
df_clif = final[final["Cool Name"] == "Cliffhanger"]
df_jet = final[final["Cool Name"] == "Jetfire"]
df_jaz = final[final["Cool Name"] == "Jazz"]
df_dino = final[final["Cool Name"] == "Dinobots"]
df_hot = final[final["Cool Name"] == "Hot Rod"]
df_opt = final[final["Cool Name"] == "Optimus Prime"]
df_sent = final[final["Cool Name"] == "Sentinel Prime"]

#See the information to see if there are missing values

#df_wheel.info()


#Putting everything in lower case for practicity and changing names for our Region

df_wheel.columns = df_wheel.columns.str.lower()
df_bumb.columns = df_bumb.columns.str.lower()
df_rat.columns = df_rat.columns.str.lower()
df_mir.columns = df_mir.columns.str.lower()
df_side.columns = df_side.columns.str.lower()
df_sm.columns = df_sm.columns.str.lower()
df_iron.columns = df_iron.columns.str.lower()
df_clif.columns = df_clif.columns.str.lower()
df_jet.columns = df_jet.columns.str.lower()
df_jaz.columns = df_jaz.columns.str.lower()
df_dino.columns = df_dino.columns.str.lower()
df_hot.columns = df_hot.columns.str.lower()
df_opt.columns = df_opt.columns.str.lower()
df_sent.columns = df_sent.columns.str.lower()

#Changing Column names

new_names = {
    'country code total'                                                       : 'country_code', 
    'country name'                                                             : 'country_name',
    'hult region'                                                              : 'hult_region', 
    'cool name'                                                                : 'cool_name', 
    'aids estimated deaths (unaids estimates)'                                 : 'est_deaths_by _aids', 
    'adjusted net enrollment rate, primary (% of primary school age children)' : 'enrollment_rate',
    'adolescent fertility rate (births per 1,000 women ages 15-19)'            : 'adolescent _fertility_rate', 
    'antiretroviral therapy coverage (% of people living with hiv)'            : 'art_therapy for aids', 
    'births attended by skilled health staff (% of total)'                     : 'births_under_obstetrician ', 
    'co2 emissions (metric tons per capita)'                                   : 'co2_emissions', 
    'contributing family workers, female (% of female employment)'             : 'female_employment_%',
    'contributing family workers, male (% of male employment)'                 : 'male_employment_%',
    'contributing family workers, total (% of total employment)'               : 'total_employment', 
    'employment to population ratio, 15+, female (%) (modeled ilo estimate)'   : 'female_employment_ratio',
    'employment to population ratio, 15+, male (%) (modeled ilo estimate)'     : 'male_employment _ratio', 
    'employment to population ratio, 15+, total (%) (modeled ilo estimate)'    : 'total_employment_ratio',
    'energy use (kg of oil equivalent) per $1,000 gdp (constant 2011 ppp)'     : 'energy_use',
    'fertility rate, total (births per woman)'                                 : 'fertility_rate', 
    'gdp per person employed (constant 2011 ppp $)'                            : 'gdp_per_person', 
    'gdp per unit of energy use (constant 2011 ppp $ per kg of oil equivalent)': 'gdp_per_energy use', 
    'gni per capita, atlas method (current US$)'                               : 'gross_national_income',
    'immunization, measles (% of children ages 12-23 months)'                  : 'measles_immunization_for kids', 
    'improved sanitation facilities (% of population with access)'             : 'improved_sanitation facilties', 
    'improved water source (% of population with access)'                      : 'improved_water resources', 
    'incidence of tuberculosis (per 100,000 people)'                           : 'tuberculosis_infections',
    'income share held by lowest 20%'                                          : 'income_share_lowest 20',
    'internet users (per 100 people)'                                          : 'internet_users',
    'life expectancy at birth, total (years)'                                  : 'life_expectancy', 
    'literacy rate, adult total (% of people ages 15 and above)'               : 'literacy_rate',
    'maternal mortality ratio (modeled estimate, per 100,000 live births)'     : 'maternal_mortality_ratio',
    'mobile cellular subscriptions (per 100 people)'                           : 'mobile_subscriptions', 
    'mortality rate, infant (per 1,000 live births)'                           : 'infant_mortality_ratio',
    'net oda received per capita (current US$)'                                : 'official_development_received',
    'population, total'                                                        : 'total_population',
    'poverty gap at national poverty lines (%)'                                : 'poverty_gap',
    'pregnant women receiving prenatal care (%)'                               : 'prenatal_care',
    'prevalence of hiv, total (% of population ages 15-49)'                    : 'hiv_prevention', 
    'prevalence of undernourishment (% of population)'                         : 'undernourishment_prevention', 
    'primary completion rate, total (% of relevant age group)'                 : 'primary_education_completion',
    'proportion of seats held by women in national parliaments (%)'            : 'women_in_parliament',
    'reported cases of malaria'                                                : 'malaria_infection',
    'school enrollment, primary (% net)'                                       : 'primary_school_enrolment', 
    'self-employed, total (% of total employment)'                             : 'self_employed', 
    'trade (% of gdp)'                                                         : 'trade', 
    'tuberculosis death rate (per 100,000 people), including hiv'              : 'death_by_tuberculosis_with aids' }

df_wheel_new = df_wheel.rename(columns=new_names)
df_bumb.rename(columns=new_names, inplace=True)
df_rat.rename(columns=new_names, inplace=True)
df_mir.rename(columns=new_names, inplace=True)
df_side.rename(columns=new_names, inplace=True)
df_sm.rename(columns=new_names, inplace=True)
df_iron.rename(columns=new_names, inplace=True)
df_clif.rename(columns=new_names, inplace=True)
df_jet.rename(columns=new_names, inplace=True)
df_jaz.rename(columns=new_names, inplace=True)
df_dino.rename(columns=new_names, inplace=True)
df_hot.rename(columns=new_names, inplace=True)
df_opt.rename(columns=new_names, inplace=True)
df_sent.rename(columns=new_names, inplace=True)

/opt/anaconda3/lib/python3.8/site-packages/pandas/core/frame.py:4295: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  return super().rename(

#Calculate the means for each region and assigning them a new name to imput later

wheel= df_wheel_new[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate']].mean(axis = 0).round(decimals = 2)

bumb = df_bumb[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate','literacy_rate']].mean(axis = 0).round(decimals = 2)

rat = df_rat[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate','literacy_rate']].mean(axis = 0).round(decimals = 2)

mir = df_mir[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate','literacy_rate']].mean(axis = 0).round(decimals = 2)

side = df_side[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate','literacy_rate']].mean(axis = 0).round(decimals = 2)

sm = df_sm[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate','literacy_rate']].mean(axis = 0).round(decimals = 2)

iron = df_iron[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate','literacy_rate']].mean(axis = 0).round(decimals = 2)

clif = df_clif[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate','literacy_rate']].mean(axis = 0).round(decimals = 2)

jet = df_jet[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate','literacy_rate']].mean(axis = 0).round(decimals = 2)

jaz = df_jaz[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate','literacy_rate']].mean(axis = 0).round(decimals = 2)

dino = df_dino[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate','literacy_rate']].mean(axis = 0).round(decimals = 2)

hot = df_hot[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate','literacy_rate']].mean(axis = 0).round(decimals = 2)

opt = df_opt[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate','literacy_rate']].mean(axis = 0).round(decimals = 2)

sent = df_sent[['women_in_parliament',
              'maternal_mortality_ratio',
              'tuberculosis_infections',
             'improved_sanitation facilties',
             'mobile_subscriptions',
             'trade',
             'gdp_per_person',
             'co2_emissions',
             'enrollment_rate','literacy_rate']].mean(axis = 0).round(decimals = 2)

#Create a Dataset with the means by region

df = pd.DataFrame({'Arabian Peninsula'                         : wheel,
                   'Western Europe'                            : jet,
                   'East/South Africa'                         : jaz,
                   'Carribean Islands & Parts of Latin America': rat,
                   'Middle East & Russia'                      : mir,
                   'Sub-Saharan Africa'                        : side,
                   'Eastern Europe'                            : sm,
                   'South America'                             : iron,
                   'Australia & Oceania'                       : clif,
                   'Central Asia'                              : dino,
                   'North America & Part of Latin America'     : hot,
                   'East/Southeast Asia'                       : opt,
                   'Northern Sub-Saharan Africa'               : bumb,})

# Transpose the DataFrame for practicity

dft = df.transpose()

Missing Value

---

In order to select the top 5 features in the data set and enhance the uniqueness when compared to the rest of the world, the strategy that we decided to use for the missing values is filled with the mean/median from our data.

For the missing values we decided to analyze the histograms of the data to see if there was any skewness to decide if the use of the mean or the median were appropriate, since we did not see any skewness, we imputed the mean to replace the missing values.

#Histogram of the mean for each 


####################Women in parliament#######################


fig, ax = plt.subplots(figsize = [10, 5])

sns.distplot(a     = dft['women_in_parliament'],
             bins  = 10,
             hist  = True,
             kde   = False,
             rug   = False,
             color = 'black')


plt.title(label   = "Distribution of women in parliament")
plt.xlabel(xlabel = 'Women in Parliament')
plt.ylabel(ylabel = 'Frequency')
plt.axvline(x     = dft['women_in_parliament'].mean(),
            color = 'red')
plt.axvline(x     = dft['women_in_parliament'].median(),
            color = 'blue')
plt.legend(labels =  ['mean', 'median'])


################### MATERNAL MORTALITY RATIO####################


fig, ax = plt.subplots(figsize = [10, 5])

sns.distplot(a     = dft['maternal_mortality_ratio'],
             bins  = 10,
             hist  = True,
             kde   = False,
             rug   = False,
             color = 'black')


plt.title(label   = "Distribution of Maternal Mortality")
plt.xlabel(xlabel = 'Maternal Mortality')
plt.ylabel(ylabel = 'Frequency')
plt.axvline(x     = dft['maternal_mortality_ratio'].mean(),
            color = 'red')
plt.axvline(x     = dft['maternal_mortality_ratio'].median(),
            color = 'blue')
plt.legend(labels =  ['mean', 'median'])



################# TUBERCULOSIS INFECTIONS #######################


fig, ax = plt.subplots(figsize = [10, 5])

sns.distplot(a     = dft['tuberculosis_infections'],
             bins  = 10,
             hist  = True,
             kde   = False,
             rug   = False,
             color = 'black')


plt.title(label   = "Distribution of Tuberculosis Infections")
plt.xlabel(xlabel = 'Tuberculosis Infections')
plt.ylabel(ylabel = 'Frequency')
plt.axvline(x     = dft['tuberculosis_infections'].mean(),
            color = 'red')
plt.axvline(x     = dft['tuberculosis_infections'].median(),
            color = 'blue')
plt.legend(labels =  ['mean', 'median'])


################# Improved Sanitation Facilities ################


fig, ax = plt.subplots(figsize = [10, 5])

sns.distplot(a     = dft['improved_sanitation facilties'],
             bins  = 10,
             hist  = True,
             kde   = False,
             rug   = False,
             color = 'black')


plt.title(label    = "Distribution of Improved Sanitation Facilities")
plt.xlabel(xlabel  = 'Improved Sanotation Facilities')
plt.ylabel(ylabel  = 'Frequency')
plt.axvline(x      = dft['improved_sanitation facilties'].mean(),
            color  = 'red')
plt.axvline(x      = dft['improved_sanitation facilties'].median(),
            color  = 'blue')
plt.legend(labels  =  ['mean', 'median'])


################# Mobile Subscriptions #########################


fig, ax = plt.subplots(figsize = [10, 5])

sns.distplot(a     = dft['mobile_subscriptions'],
             bins  = 10,
             hist  = True,
             kde   = False,
             rug   = False,
             color = 'black')


plt.title(label    = "Distribution of Mobile Subscriptions")
plt.xlabel(xlabel  = 'Mobile Subscriptions')
plt.ylabel(ylabel  = 'Frequency')
plt.axvline(x      = dft['mobile_subscriptions'].mean(),
            color  = 'red')
plt.axvline(x      = dft['mobile_subscriptions'].median(),
            color  = 'blue')
plt.legend(labels  =  ['mean', 'median'])


################### Trade ####################################


fig, ax = plt.subplots(figsize = [10, 5])

sns.distplot(a     = dft['trade'],
             bins  = 10,
             hist  = True,
             kde   = False,
             rug   = False,
             color = 'black')


plt.title(label    = "Distribution of Trade")
plt.xlabel(xlabel  = 'Trade')
plt.ylabel(ylabel  = 'Frequency')
plt.axvline(x      = dft['trade'].mean(),
            color  = 'red')
plt.axvline(x      = dft['trade'].median(),
            color  = 'blue')
plt.legend(labels  =  ['mean', 'median'])


#################### GDP per Capita ##########################



fig, ax = plt.subplots(figsize = [10, 5])

sns.distplot(a     = dft['gdp_per_person'],
             bins  = 10,
             hist  = True,
             kde   = False,
             rug   = False,
             color = 'black')


plt.title(label = "Distribution of GDP per Capita")
plt.xlabel(xlabel = 'GDP per Capita')
plt.ylabel(ylabel = 'Frequency')
plt.axvline(x = dft['gdp_per_person'].mean(),
            color = 'red')
plt.axvline(x = dft['gdp_per_person'].median(),
            color = 'blue')
plt.legend(labels =  ['mean', 'median'])

################## CO2 Emissions ###############################


fig, ax = plt.subplots(figsize = [10, 5])

sns.distplot(a     = dft['co2_emissions'],
             bins  = 10,
             hist  = True,
             kde   = False,
             rug   = False,
             color = 'black')


plt.title(label = "Distribution of CO2 Emissions")
plt.xlabel(xlabel = 'CO2 Emissions')
plt.ylabel(ylabel = 'Frequency')
plt.axvline(x = dft['co2_emissions'].mean(),
            color = 'red')
plt.axvline(x = dft['co2_emissions'].median(),
            color = 'blue')
plt.legend(labels =  ['mean', 'median'])


################# Enrollment Rate ###########################



fig, ax = plt.subplots(figsize = [10, 5])

sns.distplot(a     = dft['enrollment_rate'],
             bins  = 10,
             hist  = True,
             kde   = False,
             rug   = False,
             color = 'black')


plt.title(label = "Distribution of Enrollment Rate")
plt.xlabel(xlabel = 'Enrollment Rate')
plt.ylabel(ylabel = 'Frequency')
plt.axvline(x = dft['enrollment_rate'].mean(),
            color = 'red')
plt.axvline(x = dft['enrollment_rate'].median(),
            color = 'blue')
plt.legend(labels =  ['mean', 'median'])

/opt/anaconda3/lib/python3.8/site-packages/seaborn/distributions.py:2551: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).
  warnings.warn(msg, FutureWarning)

<matplotlib.legend.Legend at 0x123196b80>

#flagging the missing values in our Region

for col in df_wheel_new:

    if df_wheel_new[col].isnull().astype(int).sum() > 0:
        df_wheel_new['m_'+col] = df_wheel_new[col].isnull().astype(int)

#soft coding MEAN for women_in_parliament
women_mean = df_wheel_new['women_in_parliament'].mean()


#filling women_in_parliament NAs with MEDIAN
df_wheel_new['women_in_parliament'].fillna(value = women_mean,
                         inplace = True)



#soft coding MEAN for literacy_rate
literacy_mean = df_wheel_new['literacy_rate'].mean()


#filling literacy_rate NAs with MEAN
df_wheel_new['literacy_rate'].fillna(value = literacy_mean,
                         inplace = True)
#Calculating the mean for our Region for each category
wheel= df_wheel_new[['women_in_parliament',
                     'maternal_mortality_ratio',
                     'tuberculosis_infections',
                     'improved_sanitation facilties',
                     'mobile_subscriptions',
                     'trade',
                     'gdp_per_person',
                     'co2_emissions',
                     'enrollment_rate']].mean(axis = 0).round(decimals = 2)

#Calculating the mean for each Region for each category

df_bumb = final[final["Cool Name"] == "Bumblebee"]
df_rat  = final[final["Cool Name"] == "Ratchet"]
df_mir  = final[final["Cool Name"] == "Mirage"]
df_side = final[final["Cool Name"] == "Sideswipe"]
df_sm   = final[final["Cool Name"] == "Skids & Mudflap"]
df_iron = final[final["Cool Name"] == "Ironhide"]
df_clif = final[final["Cool Name"] == "Cliffhanger"]
df_jet  = final[final["Cool Name"] == "Jetfire"]
df_jaz  = final[final["Cool Name"] == "Jazz"]
df_dino = final[final["Cool Name"] == "Dinobots"]
df_hot  = final[final["Cool Name"] == "Hot Rod"]
df_opt  = final[final["Cool Name"] == "Optimus Prime"]
df_sent = final[final["Cool Name"] == "Sentinel Prime"]

new_names = {
    'country code total'                                                       : 'country_code', 
    'country name'                                                             : 'country_name',
    'hult region'                                                              : 'hult_region', 
    'cool name'                                                                : 'cool_name', 
    'aids estimated deaths (unaids estimates)'                                 : 'est_deaths_by _aids', 
    'adjusted net enrollment rate, primary (% of primary school age children)' : 'enrollment_rate',
    'adolescent fertility rate (births per 1,000 women ages 15-19)'            : 'adolescent _fertility_rate', 
    'antiretroviral therapy coverage (% of people living with hiv)'            : 'art_therapy for aids', 
    'births attended by skilled health staff (% of total)'                     : 'births_under_obstetrician ', 
    'co2 emissions (metric tons per capita)'                                   : 'co2_emissions', 
    'contributing family workers, female (% of female employment)'             : 'female_employment_%',
    'contributing family workers, male (% of male employment)'                 : 'male_employment_%',
    'contributing family workers, total (% of total employment)'               : 'total_employment', 
    'employment to population ratio, 15+, female (%) (modeled ilo estimate)'   : 'female_employment_ratio',
    'employment to population ratio, 15+, male (%) (modeled ilo estimate)'     : 'male_employment _ratio', 
    'employment to population ratio, 15+, total (%) (modeled ilo estimate)'    : 'total_employment_ratio',
    'energy use (kg of oil equivalent) per $1,000 gdp (constant 2011 ppp)'     : 'energy_use',
    'fertility rate, total (births per woman)'                                 : 'fertility_rate', 
    'gdp per person employed (constant 2011 ppp $)'                            : 'gdp_per_person', 
    'gdp per unit of energy use (constant 2011 ppp $ per kg of oil equivalent)': 'gdp_per_energy use', 
    'gni per capita, atlas method (current US$)'                               : 'gross_national_income',
    'immunization, measles (% of children ages 12-23 months)'                  : 'measles_immunization_for kids', 
    'improved sanitation facilities (% of population with access)'             : 'improved_sanitation facilties', 
    'improved water source (% of population with access)'                      : 'improved_water resources', 
    'incidence of tuberculosis (per 100,000 people)'                           : 'tuberculosis_infections',
    'income share held by lowest 20%'                                          : 'income_share_lowest 20',
    'internet users (per 100 people)'                                          : 'internet_users',
    'life expectancy at birth, total (years)'                                  : 'life_expectancy', 
    'literacy rate, adult total (% of people ages 15 and above)'               : 'literacy_rate',
    'maternal mortality ratio (modeled estimate, per 100,000 live births)'     : 'maternal_mortality_ratio',
    'mobile cellular subscriptions (per 100 people)'                           : 'mobile_subscriptions', 
    'mortality rate, infant (per 1,000 live births)'                           : 'infant_mortality_ratio',
    'net oda received per capita (current US$)'                                : 'official_development_received',
    'population, total'                                                        : 'total_population',
    'poverty gap at national poverty lines (%)'                                : 'poverty_gap',
    'pregnant women receiving prenatal care (%)'                               : 'prenatal_care',
    'prevalence of hiv, total (% of population ages 15-49)'                    : 'hiv_prevention', 
    'prevalence of undernourishment (% of population)'                         : 'undernourishment_prevention', 
    'primary completion rate, total (% of relevant age group)'                 : 'primary_education_completion',
    'proportion of seats held by women in national parliaments (%)'            : 'women_in_parliament',
    'reported cases of malaria'                                                : 'malaria_infection',
    'school enrollment, primary (% net)'                                       : 'primary_school_enrolment', 
    'self-employed, total (% of total employment)'                             : 'self_employed', 
    'trade (% of gdp)'                                                         : 'trade', 
    'tuberculosis death rate (per 100,000 people), including hiv'              : 'death_by_tuberculosis_with aids' }

#Replacing the names of the columns

#Everything in lowercase
df_bumb.columns  = df_bumb.columns.str.lower()
df_rat.columns   = df_rat.columns.str.lower()
df_mir.columns   = df_mir.columns.str.lower()
df_side.columns  = df_side.columns.str.lower()
df_sm.columns    = df_sm.columns.str.lower()
df_iron.columns  = df_iron.columns.str.lower()
df_clif.columns  = df_clif.columns.str.lower()
df_jet.columns   = df_jet.columns.str.lower()
df_jaz.columns   = df_jaz.columns.str.lower()
df_dino.columns  = df_dino.columns.str.lower()
df_hot.columns   = df_hot.columns.str.lower()
df_opt.columns   = df_opt.columns.str.lower()
df_sent.columns  = df_sent.columns.str.lower()



#The actual replacement of the names

df_bumb.rename(columns = new_names, inplace = True)
df_rat.rename(columns  = new_names, inplace = True)
df_mir.rename(columns  = new_names, inplace = True)
df_side.rename(columns = new_names, inplace = True)
df_sm.rename(columns   = new_names, inplace = True)
df_iron.rename(columns = new_names, inplace = True)
df_clif.rename(columns = new_names, inplace = True)
df_jet.rename(columns  = new_names, inplace = True)
df_jaz.rename(columns  = new_names, inplace = True)
df_dino.rename(columns = new_names, inplace = True)
df_hot.rename(columns  = new_names, inplace = True)
df_opt.rename(columns  = new_names, inplace = True)
df_sent.rename(columns = new_names, inplace = True)

/opt/anaconda3/lib/python3.8/site-packages/pandas/core/frame.py:4295: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  return super().rename(

HEATMAP

---

From the above analysis, we got these 5 features are what made our region unique, which is:

• Health-Immunization
• Life Expectancy
• Mobile Internet
• Mobile Users
• Sanitation

To understand the top features of the Arabian peninsula region, we selected the attributes like the quality of life and analyzed the associated columns like; Education enrolment, immunization, sanitation facilities, and in order to understand public life as an attribute we used associated variables like access to technology like usage of mobile devices and internet users.

We found the strongest correlation in variable sanitation facilities with respect to students’ enrollment in primary education, life expectancy, immunizations, and internet access.

The relation between sanitation other variables can be elaborated as follows:

• Educating students enrolled in primary school increases the importance of sanitation and hence implementing the appropriate facilities
  
• Higher the sanitation facilities and access to the same increases the life expectancy of an individual
  
• Reaching out to people through communication and information devices like mobile phones and the internet helps to increase the value of sanitation amenities
  

This correlation helped us in understanding how sanitation is one of the important factors to help enhance sectors like health, education, economics, and employment.

Another interesting correlation we analyzed was more the number of women participating in parliament results in higher enrolment rates of students for primary education and the other correlation was if employment rate is higher then the CO2 emissions are increased too, this can be associated with higher employment at oil refineries, factories and companies and higher use of transportations.

################################ HEAT MAP ####################################                                                                           #

wheeljack_heatmap = df_wheel_new[['country_code', 'enrollment_rate','total_employment','gdp_per_person', 
                                  'life_expectancy', 'literacy_rate','maternal_mortality_ratio',
                                  'women_in_parliament','improved_sanitation facilties',
                                  'mobile_subscriptions','internet_users','births_under_obstetrician ',
                                  'co2_emissions','measles_immunization_for kids']]

#Pearson correlation to identify the correlation between variables
wheeljack_correlation = wheeljack_heatmap.corr(method = 'pearson').round(decimals = 2)

#Plotting Heat-map
fig, ax = plt.subplots (figsize = (10,10)) 
sns.heatmap (data       = wheeljack_correlation,
             cmap       = 'viridis',
             square     = True,
             annot      = True,
             linecolor  = 'black',
             linewidths = 0.5)

plt.title ("""
Linear Correlation of the Top Features in Arabian Peninsula""")

Text(0.5, 1.0, '\nLinear Correlation of the Top Features in Arabian Peninsula')

Represent Country - Israel

---

We decided to rank countries according to the above selected variables. We attributed points (1 worst- 4 best) to the 4 first countries for each variable. We then attributed weight to each defined variable using the related SDGs (Sustainable development Goals) coverage index for the Arabian Peninsula. Finally, we summed the points for each country to obtain our best performing country, which is Israel.

ranking_wheeljack_DATA = pd.DataFrame({'country_name': ['ISRAEL', 'UNITED ARAB EMIRATES','CYPRYUS',
'OMAN','JORDAN','KUWAIT','QATAR','BAHRAIN','SAUDI ARABIA','IRAQ','LEBANON','TURKEY','YEMEN'],
                        't_employmen_rt'              : [0,2,0,0,0,1,4,3,0,0,0,0,0],
                        'gdp_person'                  : [0,0,0,1,0,4,3,0,2,0,0,0,0],
                        'gni'                         : [0,2,1,0,0,3,4,0,0,0,0,0,0], 
                        'immunization'                : [2,0,0,0,0,4,1,3,0,0,0,0,0],
                        'life_expectancy'             : [4,0,3,1,0,0,0,0,0,0,2,0,0],
                        'sanitation'                  : [3,0,1,0,0,4,0,0,2,0,0,0,0],
                        'water'                       : [4,0,3,0,0,0,1,2,0,0,0,0,0],
                        'internet_users'              : [3,4,0,0,0,0,1,2,0,0,0,0,0],
                        'mobile_subscription'         : [0,3,2,0,0,0,0,1,4,0,0,0,0],
                        'population'                  : [0,0,4,0,0,1,2,3,0,0,0,0,0],
                        'trade'                       : [0,4,1,0,2,0,0,3,0,0,0,0,0]})

for col in ranking_wheeljack_DATA:
    if col  =='t_employmen_rt' or col    =='gdp_person' or col=='gni' or col=='trade':
        ranking_wheeljack_DATA[col]=ranking_wheeljack_DATA[col]*2
    elif col=='immunization'   or col    =='life_expectancy':
        ranking_wheeljack_DATA[col]=ranking_wheeljack_DATA[col]*5
    elif col=='internet_users' or col    =='mobile_subscription':
        ranking_wheeljack_DATA[col]= ranking_wheeljack_DATA[col]*4
    elif col=='sanitation'     or col    =='water':
        ranking_wheeljack_DATA[col]= ranking_wheeljack_DATA[col]*3
    else:
        ranking_wheeljack_DATA[col]= ranking_wheeljack_DATA[col]*1

ranking_wheeljack_DATA['Total'] = ranking_wheeljack_DATA.sum(axis=1)
ranking_wheeljack_DATA["Rank"]  = ranking_wheeljack_DATA['country_name'].rank() 
ranking_wheeljack_DATA.sort_values("Total", ascending = False, inplace = True)

print("Top 5 Wheeljack region countries:")
ranking_wheeljack_DATA['country_name'].head(n=5)

Top 5 Wheeljack region countries:

0                  ISRAEL
5                  KUWAIT
7                 BAHRAIN
1    UNITED ARAB EMIRATES
2                 CYPRYUS
Name: country_name, dtype: object

Figure 3 : Israel's Flag

Figure 4 : Location of Israel

**◆ Health-Immunization**
The measles immunization rate of children who ages 12-23 months is 98%, ranking second in the Arabian Peninsula, second only to Kuwait.

**◆ Life Expectancy**
The life expectancy in Israel is 80.95, which is one of the highest in the world. The longevity factor is its streamlined and efficient medical system.

#Health-Immunization
users= df_wheel_new.groupby("country_name")["measles_immunization_for kids"].sum()
users.plot(kind="barh", fontsize=10)

<AxesSubplot:ylabel='country_name'>

#Life Expectancy 
users= df_wheel_new.groupby("country_name")["life_expectancy"].sum()
users.plot(kind="barh", fontsize=10)

<AxesSubplot:ylabel='country_name'>

**◆ Internet Users**
Looking at the data below, the Internet users (per 100 people) in Israel is 59.39, is No.2 of the top internet users in the region.

**◆ Mobile Subscription**
In the richer countries in the world, adults use smartphones in the majority, among which Israel Mobile cellular subscriptions (per 100 people) reach 126.5.

#Internet Users
users= df_wheel_new.groupby("country_name")["internet_users"].sum()
users.plot(kind="barh", fontsize=10)

<AxesSubplot:ylabel='country_name'>

#Mobile Subscription –
users= df_wheel_new.groupby("country_name")["mobile_subscriptions"].sum()
users.plot(kind="barh", fontsize=10)

<AxesSubplot:ylabel='country_name'>

**◆ Sanitation**
The Improved sanitation facilities (% of the population with access) in Israel is 100%, the highest in the world.

#Sanitation
users= df_wheel_new.groupby("country_name")["improved_sanitation facilties"].sum()
users.plot(kind="barh", fontsize=10)

<AxesSubplot:ylabel='country_name'>

The Arabian Peninsula Compare to the World

---

#the Arabian Peninsula vs Other Regions
colors = ['b', 'green', 'y', 'pink','orange','cyan','darkgrey']
#Women in Parliament
ax = dft.plot(y= 'women_in_parliament',kind='barh', color = colors, legend = False, figsize=(10,5),
                  title = 'Women in Parliament')

for bar in ax.patches:
    bar.set_facecolor('blue')

highlight = 'Arabian Peninsula'
pos = dft.index.get_loc(highlight)

ax.patches[pos].set_facecolor('green')

#Maternal Mortality Ratio
ax = dft.plot(y= 'maternal_mortality_ratio',kind='barh', color = colors, legend = False, figsize=(10,5),
                  title = 'Maternal Mortality Ratio')

for bar in ax.patches:
    bar.set_facecolor('blue')

highlight = 'Arabian Peninsula'
pos = dft.index.get_loc(highlight)

ax.patches[pos].set_facecolor('green')

#Tuberculosis Infections

ax = dft.plot(y= 'tuberculosis_infections',kind='barh', color = colors, legend = False, figsize=(10,5),
                  title = 'Tuberculosis Infections')

for bar in ax.patches:
    bar.set_facecolor('blue')

highlight = 'Arabian Peninsula'
pos = dft.index.get_loc(highlight)

ax.patches[pos].set_facecolor('green')


#Improved Sanitation Facilities

ax = dft.plot(y= 'improved_sanitation facilties',kind='barh', color = colors, legend = False, figsize=(10,5),
                  title = 'Improved Sanitation Facilities')

for bar in ax.patches:
    bar.set_facecolor('blue')

highlight = 'Arabian Peninsula'
pos = dft.index.get_loc(highlight)

ax.patches[pos].set_facecolor('green')

#Mobile Subscriptions

ax = dft.plot(y= 'mobile_subscriptions',kind='barh', color = colors, legend = False, figsize=(10,5),
                  title = 'Mobile Subscriptions')

for bar in ax.patches:
    bar.set_facecolor('blue')

highlight = 'Arabian Peninsula'
pos = dft.index.get_loc(highlight)

ax.patches[pos].set_facecolor('green')

#Trade

ax = dft.plot(y= 'trade',kind='barh', color = colors, legend = False, figsize=(10,5),
                  title = 'Trade')

for bar in ax.patches:
    bar.set_facecolor('blue')

highlight = 'Arabian Peninsula'
pos = dft.index.get_loc(highlight)

ax.patches[pos].set_facecolor('green')

#GDP per Capita

ax = dft.plot(y= 'gdp_per_person',kind='barh', color = colors, legend = False, figsize=(10,5),
                  title = 'GDP per Capita')

for bar in ax.patches:
    bar.set_facecolor('blue')

highlight = 'Arabian Peninsula'
pos = dft.index.get_loc(highlight)

ax.patches[pos].set_facecolor('green')

#CO2 Emissions

ax = dft.plot(y= 'co2_emissions',kind='barh', color = colors, legend = False, figsize=(10,5),
                  title = 'CO2 Emissions')

for bar in ax.patches:
    bar.set_facecolor('blue')

highlight = 'Arabian Peninsula'
pos = dft.index.get_loc(highlight)

ax.patches[pos].set_facecolor('green')

#Enrollment Rate

ax = dft.plot(y= 'enrollment_rate',kind='barh', color = colors, legend = False, figsize=(10,5),
                  title = 'Enrollment Rate')

for bar in ax.patches:
    bar.set_facecolor('blue')

highlight = 'Arabian Peninsula'
pos = dft.index.get_loc(highlight)

ax.patches[pos].set_facecolor('green')

**Analysis**

---

Our Analysis is based on the Sustainable Development Goals (SDG) framework of 17 goals which is set with a vision to have our world free from poverty, hunger, and health disorder. Out of the 17 goals from the SDG, we have compared and analyzed the Arabian Peninsula with the other regions and the world data. The following are the variables considered for analysis:

Health –

• Maternal Mortality Rate: In the review period, the Maternal mortality rate in Arabian Peninsula has been decreased to 52 while the world average is at 258. Countries in the Sub- Saharan region and the least developed countries in the Arabian region have the greatest number of the maternal mortality rate.
  
• Tuberculosis per 100,000 people: In Arabian Peninsula people suffer with respiratory diseases on a higher scale compared to the world, but prevention of Tuberculosis has been successfully achieved by the Arabian Peninsula region when compared to the world average.
  

Education –

• Enrolment of children for primary education: Although the average student enrolled for primary education is a bit less than 100%, the region has been successful in rethinking of the value of education making it stand much taller than the world average.
  
• Women in parliament: Although the women participation in the parliament is lower when compared to world data but the law to involve women in the parliament has helped to increase women participation in the parliament which has improved in creating a balance in the political, economic and social life.
  

Digitization –

• Mobile subscribers: Enhanced use of primary introduction of technology for communication and information has increased in the Arabian Peninsula. This has helped the region to have higher proportion of individuals who own the devices compared to the world average.
  

Sanitation –

• While the development in providing basic sanitation varies from country to country in the Arabian Peninsula, the average proportion of population having access to these facilities is still higher than the world average.
  

Trade –

• Petroleum plays an important part in the total economic and industrial development of this region. Arabian Peninsula is the world’s largest producer and exporter of oil and in recent times this region has also diversified the exports of other industrial goods making the region’s GDP better as compared to world average.

Conclusion

---

The Arabian Peninsula in recent years is on track to attain the development goals when compared to world standards. The key finding from the data analysis are as follows -

• Gender inequality : Although the Arabian countries have made significant progress in providing equal rights to women, the region still lacks behind when compared to the world. It would be essential for this region to overcome their cultural, social, and structural roadblock in order to make more economic and social development.
  
  
• Employment and Economic growth : The development in terms of GDP has been steady but slow-paced. The oil-rich countries concentrate more on gaining higher revenue throuh oil production and exports, while other sectors are heavily dependent on oil. Other occupations like agriculture have lesser share but have a higher employment rate. Hence, it is necessary to diversify their means of increasing the economy and not be heavily dependent on oil.
  
  
• Climate Change : The emissions per capita for the Arabian Peninsula are significantly higher compared to the world average. The region suffers from climate change in a way by experiencing drought conditions, temperature rise, and so on. Hence in order to take the emissions under control adequate policies and planning should be executed at a larger capacity throughout the region.

Figure 5 : Relief Map of the Arabian Peninsula

References

---

• Countries in Arabian Peninsula - https://www.worldatlas.com/articles/which-countries-are-located-in-the-arabian-peninsula.html
  
• Visualization of GDP, Population, etc. https://data.worldbank.org/?locations=1A-ZQ-PS
  
• Agriculture - https://gro-intelligence.com/insights/articles/agricultural-technology-in-the-arabian-peninsula
  
• Groundwater resources - https://eos.org/articles/arid-arabian-peninsula-is-tapping-into-vast-groundwater-reserves
  
• Population Diversity - https://www.nature.com/articles/s41598-020-72283-1
  
• Health and Diseases - https://www.clinicalmicrobiologyandinfection.com/article/S1198-743X(14)60742-4/fulltext
  
• Arab Sustainable Development Report 2020 - https://asdr.unescwa.org/sdgs/pdf/en/ASDR2020-Final-Online.pdf
  
• Economy & Global Trade - https://www.saudiembassy.net/economy-global-trade
  
• Timeliness and completeness of routine childhood vaccinations - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6373067/
  
• Here’s why, despite it all, Israelis live longer - https://blogs.timesofisrael.com/heres-why-despite-it-all-israelis-live-longer/
  
• Smartphone Ownership Is Growing Rapidly Around the World - https://www.pewresearch.org/global/2019/02/05/smartphone-ownership-is-growing-rapidly-around-the-world-but-not-always-equally/
  
• Global survey ranks Israel second highest in smartphone usage _ https://www.i24news.tv/en/news/israel/society/194712-190205-global-survey-ranks-israel-second-highest-in-smartphone-usage
  
• #News Israelis are world’s top smartphone users - https://www.israel21c.org/israelis-are-worlds-top-smartphone-users/
  

Figure 6: the Arabian Peninsula