This work extracted open source data from World Bank focusing on the study about relationship between fertility rate, public expenditure on education and GDP. Data is stored in MySQL database for the convenience of inter-working between Python and R. Visualization is done with ggplot2 package in R.

ggplot of World Bank Data

#### Author: Bohan Zhang | The Business Analytics Program of the George Washington University
#### R part

library("RMySQL")
library("ggplot2")

con <- dbConnect(MySQL(), user="root", password="",
                  dbname="DB", host="localhost")

mydata <- dbReadTable(con, "mydata")

# Bubble Chart for Expenditure on Education, Fertility Rate and GDP per Capita
ggplot(mydata, aes(x=Expenditure_on_Education, y=Fertility_Rate, size=GDP, label=Country))+
  geom_point(colour="white", fill="red", shape=21)+ 
  scale_size_area(breaks=c(250, 500, 1000, 10000, 40000), "GDP per capita\n(constant 2005 US$)",max_size = 25)+
  stat_smooth(method="lm", size=0.5, colour="black", alpha=0.4, level=0.95)+
  scale_x_continuous(name="Public spending on education, total (% of GDP)", limits=c(0,8))+
  scale_y_continuous(name="Fertility rate, total (births per woman)", limits=c(1,7))+
  geom_text(size=4)+
  ggtitle("Fertility Rate v.s. Expenditure on Education") +
  theme_bw()

# Bar Chart for Fertility rate and Region, sorted by Fertility rate
ggplot(data=mydata, aes(reorder(factor(Country),-Fertility_Rate),y=Fertility_Rate, fill=Region)) + geom_bar(stat="identity")+
  scale_x_discrete(name='Country') + ggtitle("Fertility Rate by Region") + 
  theme(title = element_text(size=rel(1.2)), axis.text.x = element_text(size = rel(1.8), angle = 90, hjust = 1))

# Bar chart for Fertility rate and income level, sorted by Fertility rate
ggplot(data=mydata, aes(reorder(factor(Country),-Fertility_Rate),y=Fertility_Rate, fill=Income_Level)) + geom_bar(stat="identity")+
  scale_x_discrete(name='Country') + ggtitle("Fertility Rate by Income Level") + 
  theme(title = element_text(size = rel(1.2)), axis.text.x = element_text(size = rel(1.8), angle = 90, hjust = 1))

Working with World Bank API

# -*- coding: utf-8 -*-
#### Author: Bohan Zhang | The Business Analytics Program of the George Washington University
#### Python Part

import wbdata
import pandas as pd
import datetime
import MySQLdb as myDB

#### test if data for certain indicator, country and year is available
def test_year(ind, ctry, year):
    import wbdata    
    data = []    
    #### limit time frame to one year as the study only need one year data    
    data_date = (datetime.datetime(year, 1, 1), datetime.datetime(year, 1, 1))
    data = wbdata.get_data(ind, ctry, data_date)
    for c in data:
        if c['value'] == None:
            print 'Country:', c['country']['value']
            print 'Data:', c['indicator']['value']
            print 'Year:', c['date']
            print 'Staus:', 'NOT available'
        else:
            print 'Country:', c['country']['value']
            print 'Data:', c['indicator']['value']
            print 'Year:', c['date']            
            print 'Staus:', 'Available'
    
#### extract data from returned dictionary
def get_value(ind, ctry, year):
    raw_data = []
    data_date = (datetime.datetime(year, 1, 1), datetime.datetime(year, 1, 1))
    raw_data = wbdata.get_data(ind, country = ctry, data_date=data_date)
  
    data = []
    for country in raw_data:
        data.append(country['value'])
    return data
    
#### extract country names from returned dictionary
def get_countryList(ind, ctry, year):
    raw_data = []
    data_date = (datetime.datetime(year, 1, 1), datetime.datetime(year, 1, 1))
    raw_data = wbdata.get_data(ind, country = ctry, data_date=data_date)    
    
    data = []    
    for country in raw_data:
        data.append(country['country']['value'])
    return data
    
countries=["AGO","BEN","GMB","ZMB","SAU","IND","EGY","IRN","USA",
           "GBR","FRA","MLI","AUS","COL","NPL","BTN","ALB","TGO",
           "SEN","TZA","NZL","PER","LBN","PAK","PHL","IDN","THA",
           "OMN","ISR","BRA","CHL","GIN","SLE"]

#### convert string type to float type for numeric data
GDP = [float(x) for x in get_value("NY.GDP.PCAP.KD",countries,2005)]
Exp_Edu = [float(x) for x in get_value("SE.XPD.TOTL.GD.ZS",countries,2005)]
Fer_Rate = [float(x) for x in get_value("SP.DYN.TFRT.IN",countries,2005)]
Country = get_countryList("NY.GDP.PCAP.KD",countries,2005)

#### get region information for given country
def get_region(countries):
    regions = []
    all = wbdata.get_country(countries)    
    for country in all:
        region = country['region']['value']
        #### cut residual information        
        flag = region.find('(')
        if flag>0:        
            region = region[0:flag-1]
        regions.append(region)
    return regions
    
#### get income level information for given country
def get_incomeLevel(countries):
    incomes = []
    all = wbdata.get_country(countries)
    for country in all:
        income = country['incomeLevel']['value']
        #### cut residual information          
        flag = income.find(':')
        if flag>0:
            income = income[0:flag]
        incomes.append(income)
    return incomes
        
Region = get_region(countries)
Income_Level = get_incomeLevel(countries)
        
#### create a dictionary for data frame
myDic = { 'Country':  Country,
          'GDP':      GDP,
          'Expenditure on Education': Exp_Edu,
          'Fertility Rate': Fer_Rate,
          'Income Level': Income_Level,
          'Region': Region
        }

#### convert dictionary to data frame
myDF = pd.DataFrame(myDic)

#### create connection to MySQL
conn = myDB.connect('localhost','root')
cursor = conn.cursor()

sql = ' SHOW DATABASES; '
cursor.execute(sql)

#### test if the database name is used before, drop if used
sql = ' DROP DATABASE IF EXISTS DB; '
cursor.execute(sql)

sql = ' CREATE DATABASE DB; '
cursor.execute(sql)

#### create connection to the database created in last step
mydb = myDB.connect(host='localhost', user='root',passwd='', db='DB')

#### upload data frame to MySQL
myDF.to_sql(con = mydb, name = 'mydata', if_exists = 'replace', flavor = 'mysql')