Evolution of the Carbon Dioxide Emissions over Years

Header photo by Carlos “Grury” Santos

---

# needed libs 
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import folium
import requests

import warnings
warnings.simplefilter(action='ignore', category=FutureWarning)
pd.set_option('display.max_columns', 100)

Introduction

Carbon dioxide (chemical formula CO 2) […] is the most significant long-lived greenhouse gas in Earth’s atmosphere. Since the Industrial Revolution anthropogenic emissions – primarily from use of fossil fuels and deforestation – have rapidly increased its concentration in the atmosphere, leading to global warming […].

Source - Wikipedia

Goal: This analysis will show the evolution of CO 2 emissions over the last decades. It’s organized in two steps : first per capita, then for each entire country. So what are the countries with the highest emissions ?

---

Different datasets aggregation

Informations per capita

The dataset CO2_per_capita.csv comes from the github repo of Cabonmap for more infos on where the data come from, please visite their website and graphics which are very instructives. An other dataset can be found here

# Load the CSV file  / ParserError: Error tokenizing data. C error: Expected 1 fields in line 1094, saw 2 // with no delimiter
df = pd.read_csv('input/CO2_per_capita.csv', delimiter=';')
df.head()

	Country Name	Country Code	Year	CO2 Per Capita (metric tons)
0	Aruba	ABW	1960	NaN
1	Aruba	ABW	1961	NaN
2	Aruba	ABW	1962	NaN
3	Aruba	ABW	1963	NaN
4	Aruba	ABW	1964	NaN

Columns names are self explanatory.

df.Year.unique()

array([1960, 1961, 1962, 1963, 1964, 1965, 1966, 1967, 1968, 1969, 1970,
       1971, 1972, 1973, 1974, 1975, 1976, 1977, 1978, 1979, 1980, 1981,
       1982, 1983, 1984, 1985, 1986, 1987, 1988, 1989, 1990, 1991, 1992,
       1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003,
       2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011])

Country codes and continents

This dataset consists of list of countries by continent. Continent codes and country codes are also included. Credits : JohnSnowLabs via Datahub.io

df_continent = pd.read_csv('input/country-and-continent-codes-list-csv_csv.csv')
df_continent.head()

	Continent_Name	Continent_Code	Country_Name	Two_Letter_Country_Code	Three_Letter_Country_Code	Country_Number
0	Asia	AS	Afghanistan, Islamic Republic of	AF	AFG	4.0
1	Europe	EU	Albania, Republic of	AL	ALB	8.0
2	Antarctica	AN	Antarctica (the territory South of 60 deg S)	AQ	ATA	10.0
3	Africa	AF	Algeria, People's Democratic Republic of	DZ	DZA	12.0
4	Oceania	OC	American Samoa	AS	ASM	16.0

# select only interesting cols
df_continent = df_continent[['Continent_Name', 'Three_Letter_Country_Code']]
# rename them
df_continent.columns = ['Continent', 'Country Code']
# merge two df
df = pd.merge(df, df_continent, on='Country Code')
df.head()

	Country Name	Country Code	Year	CO2 Per Capita (metric tons)	Continent
0	Aruba	ABW	1960	NaN	North America
1	Aruba	ABW	1961	NaN	North America
2	Aruba	ABW	1962	NaN	North America
3	Aruba	ABW	1963	NaN	North America
4	Aruba	ABW	1964	NaN	North America

df_continent.shape

(262, 2)

df_continent.isnull().sum()

Continent       0
Country Code    4
dtype: int64

Countries population over years

This database presents population and other demographic estimates and projections from 1960 to 2050. They are disaggregated by age-group and sex and covers more than 200 economies. Here i’ll keep only relevant infos for our analysis. The db come from worldbank.org

df_population = pd.read_csv('input/Population-EstimatesData.csv')

# keep only total population
df_population = df_population[df_population['Indicator Name'] == 'Population, total']

# keep only corresponding years and remove unecessary cols
df_population = df_population.drop(columns=['Country Name', 'Indicator Name', 'Indicator Code', '2012', '2013',
       '2014', '2015', '2016', '2017', '2018', '2019', '2020', '2021', '2022',
       '2023', '2024', '2025', '2026', '2027', '2028', '2029', '2030', '2031',
       '2032', '2033', '2034', '2035', '2036', '2037', '2038', '2039', '2040',
       '2041', '2042', '2043', '2044', '2045', '2046', '2047', '2048', '2049',
       '2050', 'Unnamed: 95'])
df_population.head()

	Country Code	1960	1961	1962	1963	1964	1965	1966	1967	1968	1969	1970	1971	1972	1973	1974	1975	1976	1977	1978	1979	1980	1981	1982	1983	1984	1985	1986	1987	1988	1989	1990	1991	1992	1993	1994	1995	1996	1997	1998	1999	2000	2001	2002	2003	2004	2005	2006	2007	2008	2009	2010	2011
166	ARB	9.249093e+07	9.504450e+07	9.768229e+07	1.004111e+08	1.032399e+08	1.061750e+08	1.092306e+08	1.124069e+08	1.156802e+08	1.190165e+08	1.223984e+08	1.258074e+08	1.292694e+08	1.328634e+08	1.366968e+08	1.408433e+08	1.453324e+08	1.501331e+08	1.551837e+08	1.603925e+08	1.656895e+08	1.710520e+08	1.764901e+08	1.820058e+08	1.876108e+08	1.933103e+08	1.990938e+08	2.049425e+08	2.108448e+08	2.167874e+08	2.247354e+08	2.308299e+08	2.350372e+08	2.412861e+08	2.474359e+08	2.550297e+08	2.608435e+08	2.665751e+08	2.722351e+08	2.779629e+08	2.838320e+08	2.898504e+08	2.960266e+08	3.024345e+08	3.091620e+08	3.162647e+08	3.237733e+08	3.316538e+08	3.398255e+08	3.481451e+08	3.565089e+08	3.648959e+08
341	CSS	4.198307e+06	4.277802e+06	4.357746e+06	4.436804e+06	4.513246e+06	4.585777e+06	4.653919e+06	4.718167e+06	4.779624e+06	4.839881e+06	4.900059e+06	4.960647e+06	5.021359e+06	5.082049e+06	5.142246e+06	5.201705e+06	5.260062e+06	5.317542e+06	5.375393e+06	5.435143e+06	5.497756e+06	5.564200e+06	5.633661e+06	5.702754e+06	5.766957e+06	5.823242e+06	5.870023e+06	5.908886e+06	5.943661e+06	5.979907e+06	6.021614e+06	6.070204e+06	6.124265e+06	6.181538e+06	6.238576e+06	6.292827e+06	6.343683e+06	6.392040e+06	6.438587e+06	6.484510e+06	6.530691e+06	6.577216e+06	6.623792e+06	6.670276e+06	6.716373e+06	6.761932e+06	6.806838e+06	6.851221e+06	6.895315e+06	6.939534e+06	6.984096e+06	7.029022e+06
516	CEB	9.140176e+07	9.223274e+07	9.300950e+07	9.384002e+07	9.471580e+07	9.544099e+07	9.614634e+07	9.704327e+07	9.788402e+07	9.860663e+07	9.913455e+07	9.963526e+07	1.003572e+08	1.011127e+08	1.019399e+08	1.028606e+08	1.037761e+08	1.046169e+08	1.053294e+08	1.059486e+08	1.065767e+08	1.071915e+08	1.077700e+08	1.083261e+08	1.088535e+08	1.093607e+08	1.098466e+08	1.102964e+08	1.106867e+08	1.108016e+08	1.107431e+08	1.104695e+08	1.101115e+08	1.100419e+08	1.100216e+08	1.098642e+08	1.096262e+08	1.094220e+08	1.092383e+08	1.090610e+08	1.084478e+08	1.076600e+08	1.069598e+08	1.066242e+08	1.063317e+08	1.060419e+08	1.057725e+08	1.053787e+08	1.050019e+08	1.048005e+08	1.044214e+08	1.041740e+08
691	EAR	9.792874e+08	1.002524e+09	1.026587e+09	1.051415e+09	1.077037e+09	1.103433e+09	1.130587e+09	1.158571e+09	1.187274e+09	1.216766e+09	1.247053e+09	1.278138e+09	1.310016e+09	1.342709e+09	1.376073e+09	1.410094e+09	1.444720e+09	1.480010e+09	1.516216e+09	1.553704e+09	1.592674e+09	1.633180e+09	1.675079e+09	1.718098e+09	1.761829e+09	1.805996e+09	1.850487e+09	1.895290e+09	1.940220e+09	1.985084e+09	2.031828e+09	2.076398e+09	2.120567e+09	2.164508e+09	2.208444e+09	2.252579e+09	2.297015e+09	2.341634e+09	2.386185e+09	2.430487e+09	2.474601e+09	2.518353e+09	2.561813e+09	2.605067e+09	2.648272e+09	2.691528e+09	2.734860e+09	2.778276e+09	2.821797e+09	2.865440e+09	2.909411e+09	2.953406e+09
866	EAS	1.040034e+09	1.043597e+09	1.058046e+09	1.083797e+09	1.109192e+09	1.135651e+09	1.165546e+09	1.194209e+09	1.223467e+09	1.256390e+09	1.289320e+09	1.323021e+09	1.354873e+09	1.385130e+09	1.415205e+09	1.442315e+09	1.466537e+09	1.489432e+09	1.512228e+09	1.535457e+09	1.558242e+09	1.581867e+09	1.607789e+09	1.633686e+09	1.658311e+09	1.683505e+09	1.710226e+09	1.738329e+09	1.766707e+09	1.794458e+09	1.821518e+09	1.847580e+09	1.871877e+09	1.895331e+09	1.918823e+09	1.941909e+09	1.964618e+09	1.986766e+09	2.008138e+09	2.028093e+09	2.047139e+09	2.065520e+09	2.082948e+09	2.099537e+09	2.115551e+09	2.131356e+09	2.147021e+09	2.162088e+09	2.177418e+09	2.192343e+09	2.207155e+09	2.221935e+09

df_population.shape

(259, 53)

There are many missing value here, so a little cleaning is needed first

#df_population.isnull().sum()
#df_population[df_population['1960'].isnull()]

df_population = df_population.drop(index=5066)

cols_with_nan = ['1960', '1961', '1962', '1963', '1964', 
    '1965', '1966', '1967', '1968', '1969', '1970', '1971', '1972', '1973', '1974', '1975', '1976', '1977', 
    '1978', '1979', '1980', '1981', '1982', '1983', '1984','1985', '1986', '1987', '1988', '1989']
idx = [36916, 44791]
    
df_population.loc[idx, cols_with_nan] = df_population.loc[idx, '1990']

df_population.loc[37616] = df_population.loc[37616].fillna(df_population.loc[37616, '1998'])

df_population = df_population.melt(id_vars=["Country Code"], 
            #value_vars :  Column(s) to unpivot. If not specified, uses all columns that are not set as `id_vars`. 
            value_name="Population")

# Create a unique key for future join
#df_population['key'] = df_population['Country Code'] + str(df_population['variable'])
#df_population.head()

df_population = df_population.rename(index=str, columns={"variable": "Year"})
df_population.Year = df_population.Year.astype('int')
df_population.head()

	Country Code	Year	Population
0	ARB	1960	9.249093e+07
1	CSS	1960	4.198307e+06
2	CEB	1960	9.140176e+07
3	EAR	1960	9.792874e+08
4	EAS	1960	1.040034e+09

Aggregation of all datasets

df = pd.merge(df, df_population, on=['Country Code', 'Year'])
df.head()

	Country Name	Country Code	Year	CO2 Per Capita (metric tons)	Continent	Population
0	Aruba	ABW	1960	NaN	North America	54211.0
1	Aruba	ABW	1961	NaN	North America	55438.0
2	Aruba	ABW	1962	NaN	North America	56225.0
3	Aruba	ABW	1963	NaN	North America	56695.0
4	Aruba	ABW	1964	NaN	North America	57032.0

# let's check values
#temp[temp['Country Name'] == 'France']

First insights / data cleaning

Number of lines, types of values, irrelevant or weird values…

df.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 11388 entries, 0 to 11387
Data columns (total 6 columns):
Country Name                    11388 non-null object
Country Code                    11388 non-null object
Year                            11388 non-null int64
CO2 Per Capita (metric tons)    9233 non-null float64
Continent                       11388 non-null object
Population                      11385 non-null float64
dtypes: float64(2), int64(1), object(3)
memory usage: 622.8+ KB

df.shape

(11388, 6)

df.duplicated().sum()

0

df.loc[[2650, 2651, 10502, 10503]]

	Country Name	Country Code	Year	CO2 Per Capita (metric tons)	Continent	Population
2650	Cyprus	CYP	2011	6.735376	Europe	1124835.0
2651	Cyprus	CYP	2011	6.735376	Asia	1124835.0
10502	Turkey	TUR	2011	4.383105	Europe	73409455.0
10503	Turkey	TUR	2011	4.383105	Asia	73409455.0

df = df.drop(index=[2651, 10503])
df.shape

(11386, 6)

df.isnull().sum()

Country Name                       0
Country Code                       0
Year                               0
CO2 Per Capita (metric tons)    2155
Continent                          0
Population                         3
dtype: int64

# Nb of different countries
df['Country Name'].nunique()

212

# Nb of years
df['Year'].nunique()

52

plt.figure(figsize=(16, 6))
sns.distplot(df['CO2 Per Capita (metric tons)'].dropna())
# same thing but longer
#sns.distplot(df[df['CO2 Per Capita (metric tons)'].notnull()]['CO2 Per Capita (metric tons)'])
plt.show()

• At first glance, there are many years/countries with little emissions while very few countries seem to produce a lot of CO2… Let’s check this later with other plots.
• There is not any abnormal negative values. Now, where are the missing values i.e in which countries are there only missing values ? What is the proportion of Nan per country…

df[df['CO2 Per Capita (metric tons)'].isna()]['Year'].value_counts().plot(kind='bar', figsize=(16, 6))

<matplotlib.axes._subplots.AxesSubplot at 0x7fa36b8a89b0>

It seems that emissions were not fully recorded before the 90’s… Let’s dig a little deeper.

# Countries by number of missing values - there are 52 years in the record
df[df['CO2 Per Capita (metric tons)'].isna()]['Country Name'].value_counts().plot(kind='bar', figsize=(16, 6))

<matplotlib.axes._subplots.AxesSubplot at 0x7fa36b774898>

On the bar plot above, one can see that

• exept Ukraine, Russia, Croatia, Germany
• countries with at least 20 missing values for 52 years of record are not big countries.

Therefore, they can be omitted in our analysis.

# retrieve countries with a least 20 years of missing values
temp_df = df[df['CO2 Per Capita (metric tons)'].isna()]['Country Name'].value_counts() > 20
countries_with_na = pd.DataFrame(temp_df).index
countries_with_na

Index(['Armenia', 'Kazakhstan', 'Azerbaijan', 'Russian Federation', 'Georgia',
       'San Marino', 'South Sudan', 'Sint Maarten (Dutch part)',
       'Virgin Islands (U.S.)', 'Isle of Man', 'Guam', 'Curacao', 'Monaco',
       'St. Martin (French part)', 'American Samoa', 'Puerto Rico',
       'Northern Mariana Islands', 'Tuvalu', 'Liechtenstein', 'Serbia',
       'Montenegro', 'Lesotho', 'Timor-Leste', 'Korea, Dem. People_s Rep.',
       'West Bank and Gaza', 'Micronesia, Fed. Sts.', 'Andorra',
       'Turks and Caicos Islands', 'Eritrea', 'Moldova', 'Macedonia, FYR',
       'Czech Republic', 'Croatia', 'Slovenia', 'Latvia', 'Uzbekistan',
       'Kyrgyz Republic', 'Estonia', 'Belarus', 'Tajikistan',
       'Slovak Republic', 'Turkmenistan', 'Lithuania', 'Ukraine',
       'Bosnia and Herzegovina', 'Germany', 'Marshall Islands', 'Namibia',
       'Aruba', 'Bangladesh', 'Botswana', 'Maldives', 'Malaysia', 'Bhutan',
       'Oman', 'Zambia', 'Somalia', 'Zimbabwe', 'Malawi', 'Seychelles',
       'Kuwait', 'Vanuatu', 'Swaziland', 'Burundi', 'Kiribati', 'Senegal'],
      dtype='object')

# removing countries with more than 20 missing values
df = df[~df['Country Name'].isin(countries_with_na)]
df.shape

(7694, 6)

# filling remaining missing values with an interpolation 
df = df.interpolate()

# check if there isn't any Nan anymore
df.isnull().sum()

Country Name                    0
Country Code                    0
Year                            0
CO2 Per Capita (metric tons)    0
Continent                       0
Population                      0
dtype: int64

---

Analysis per capita

Which countries have the highest emissions historically ?

df_hist = pd.DataFrame(df.groupby(by='Country Name', as_index=False)['CO2 Per Capita (metric tons)'].mean())
df_hist = df_hist.sort_values(by=['CO2 Per Capita (metric tons)'], ascending=False)
df_hist.head()

	Country Name	CO2 Per Capita (metric tons)
111	Qatar	54.423341
139	United Arab Emirates	31.844877
82	Luxembourg	28.196509
17	Brunei Darussalam	21.497854
9	Bahrain	19.867874

sns.set(style="whitegrid")
plt.figure(figsize=(16, 40))
sns.barplot(x="CO2 Per Capita (metric tons)", 
            y="Country Name", 
            data=df_hist)
plt.show()

Which countries have the highest emissions lately ?

# for instance in year 2011
df_lately = df[df.Year == 2011]
df_lately = df_lately.sort_values(by=['CO2 Per Capita (metric tons)'], ascending=False)
plt.figure(figsize=(16, 40))
ax = sns.barplot(x="CO2 Per Capita (metric tons)", y="Country Name", data=df_lately)

Are the annual emissions decreasing or increasing ?

Let’s select few countries to show the evolution

selected_countries = ['France', 'Israel', 'Switzerland', 'Chile', 'China', 
                      'Colombia', 'United Kingdom', 'United States', 'Brazil', 'Australia']
plt.figure(figsize=(12, 8))
sns.lineplot(x="Year", 
             y="CO2 Per Capita (metric tons)", 
             hue="Country Name", 
             data=df[df["Country Name"].isin(selected_countries)])
plt.title('Evolution of the CO2 emissions per capita for few countries')
plt.show()

plt.figure(figsize=(12, 8))
plt.title('Evolution of the emissions per capita for each continent')

sns.lineplot(x="Year", 
             y="CO2 Per Capita (metric tons)", 
             hue="Continent", 
             data=df)

plt.show()

df_mean = pd.DataFrame(df.groupby(by=['Continent', 'Year'], as_index=False)['CO2 Per Capita (metric tons)'].mean())

plt.figure(figsize=(12, 8))
plt.title('Evolution of the MEAN emissions per capita for each continent')

sns.lineplot(x="Year", 
             y="CO2 Per Capita (metric tons)", 
             hue="Continent", 
             data=df_mean)

plt.show()

df_mean.head()

	Continent	Year	CO2 Per Capita (metric tons)
0	Africa	1960	0.298993
1	Africa	1961	0.310182
2	Africa	1962	0.308247
3	Africa	1963	0.323735
4	Africa	1964	0.348756

Evolution of emission share

df_mean_pivot = pd.pivot_table(df_mean, index='Year', values='CO2 Per Capita (metric tons)', columns='Continent')
df_mean_pivot.head()

Continent	Africa	Asia	Europe	North America	Oceania	South America
Year
1960	0.298993	1.015958	5.310022	2.134758	2.734202	1.563707
1961	0.310182	1.230886	5.445797	2.348035	3.131762	1.508886
1962	0.308247	1.290272	5.703817	2.543537	2.683807	1.549366
1963	0.323735	4.151895	5.997686	2.315408	2.848258	1.549151
1964	0.348756	4.054674	6.281629	2.664622	3.515654	1.595202

df_mean_perc = df_mean_pivot.divide(df_mean_pivot.sum(axis=1), axis=0)

plt.figure(figsize=(12, 8))

# Make the plot
plt.stackplot(range(1,53),
              df_mean_perc['Africa'], 
              df_mean_perc["Asia"], 
              df_mean_perc["Europe"],
              df_mean_perc["North America"],
              df_mean_perc["Oceania"],
              df_mean_perc["South America"],
              labels=['Africa','Asia','Europe','North America','Oceania','South America'])

# Formatting the plot
plt.legend(loc='upper left')
plt.margins(0,0)
plt.title('Evolution of emissions per capita share over time')
plt.show()

World map

# create a map
m = folium.Map()

countries_list = list(df["Country Name"].unique())

# removing names that are not recognized by the API
rem = ['Congo, Dem. Rep.', 'Congo, Rep.', 'Egypt, Arab Rep.', 'French Polynesia', 'Hong Kong SAR, China',
      'Iran, Islamic Rep.', 'Korea, Rep.', 'Lao PDR', 'Macao SAR, China', 'New Caledonia', 'Philippines', 
       'Venezuela, RB', 'Yemen, Rep.']

for c in rem:
    countries_list.remove(c)
    
countries_list.sort()
countries_list

['Afghanistan',
 'Albania',
 'Algeria',
 'Angola',
 'Antigua and Barbuda',
 'Argentina',
 'Australia',
 'Austria',
 'Bahamas, The',
 'Bahrain',
 'Barbados',
 'Belgium',
 'Belize',
 'Benin',
 'Bermuda',
 'Bolivia',
 'Brazil',
 'Brunei Darussalam',
 'Bulgaria',
 'Burkina Faso',
 'Cabo Verde',
 'Cambodia',
 'Cameroon',
 'Canada',
 'Cayman Islands',
 'Central African Republic',
 'Chad',
 'Chile',
 'China',
 'Colombia',
 'Comoros',
 'Costa Rica',
 "Cote d'Ivoire",
 'Cuba',
 'Cyprus',
 'Denmark',
 'Djibouti',
 'Dominica',
 'Dominican Republic',
 'Ecuador',
 'El Salvador',
 'Equatorial Guinea',
 'Ethiopia',
 'Faroe Islands',
 'Fiji',
 'Finland',
 'France',
 'Gabon',
 'Gambia, The',
 'Ghana',
 'Greece',
 'Greenland',
 'Grenada',
 'Guatemala',
 'Guinea',
 'Guinea-Bissau',
 'Guyana',
 'Haiti',
 'Honduras',
 'Hungary',
 'Iceland',
 'India',
 'Indonesia',
 'Iraq',
 'Ireland',
 'Israel',
 'Italy',
 'Jamaica',
 'Japan',
 'Jordan',
 'Kenya',
 'Lebanon',
 'Liberia',
 'Libya',
 'Luxembourg',
 'Madagascar',
 'Mali',
 'Malta',
 'Mauritania',
 'Mauritius',
 'Mexico',
 'Mongolia',
 'Morocco',
 'Mozambique',
 'Myanmar',
 'Nepal',
 'Netherlands',
 'New Zealand',
 'Nicaragua',
 'Niger',
 'Nigeria',
 'Norway',
 'Pakistan',
 'Palau',
 'Panama',
 'Papua New Guinea',
 'Paraguay',
 'Peru',
 'Poland',
 'Portugal',
 'Qatar',
 'Romania',
 'Rwanda',
 'Samoa',
 'Sao Tome and Principe',
 'Saudi Arabia',
 'Sierra Leone',
 'Singapore',
 'Solomon Islands',
 'South Africa',
 'Spain',
 'Sri Lanka',
 'St. Kitts and Nevis',
 'St. Lucia',
 'St. Vincent and the Grenadines',
 'Sudan',
 'Suriname',
 'Sweden',
 'Switzerland',
 'Syrian Arab Republic',
 'Tanzania',
 'Thailand',
 'Togo',
 'Tonga',
 'Trinidad and Tobago',
 'Tunisia',
 'Turkey',
 'Uganda',
 'United Arab Emirates',
 'United Kingdom',
 'United States',
 'Uruguay',
 'Vietnam']

def get_boundingbox_country(country, output_as='boundingbox'):
    """
    get the bounding box of a country in EPSG4326 given a country name

    Parameters
    ----------
    country : str
        name of the country in english and lowercase
    output_as : 'str
        chose from 'boundingbox' or 'center'. 
         - 'boundingbox' for [latmin, latmax, lonmin, lonmax]
         - 'center' for [latcenter, loncenter]

    Returns
    -------
    output : list
        list with coordinates as str
    """
    # create url
    url = '{0}{1}{2}'.format('http://nominatim.openstreetmap.org/search?country=',
                             country,
                             '&format=json&polygon=0')
    response = requests.get(url).json()[0]

    # parse response to list
    if output_as == 'boundingbox':
        lst = response[output_as]
        output = [float(i) for i in lst]
    if output_as == 'center':
        lst = [response.get(key) for key in ['lat','lon']]
        output = [float(i) for i in lst]
    return output

# Example
print("Coordinates of France are long={} and lat={}".format(
            get_boundingbox_country("El Salvador", output_as="center")[0],
            get_boundingbox_country("El Salvador", output_as="center")[1]))

Coordinates of France are long=13.8000382 and lat=-88.9140683

df_lately[df_lately['Country Name'] == 'Turkey']['CO2 Per Capita (metric tons)']

10502    4.383105
Name: CO2 Per Capita (metric tons), dtype: float64

for country in countries_list:
    resp = get_boundingbox_country(country)
    long, lat = resp[0], resp[1]
    emission = float(df_lately[df_lately['Country Name'] == country]['CO2 Per Capita (metric tons)'].values)
    folium.Circle(
                location=[long, lat],
                popup=country,
                radius=100 * emission
                ).add_to(m)

m

Analysis per whole country emission

Which countries have the highest emissions historically ?

df['Whole emission'] = df['CO2 Per Capita (metric tons)'] * df['Population']
df.head()

	Country Name	Country Code	Year	CO2 Per Capita (metric tons)	Continent	Population	Whole emission
52	Afghanistan	AFG	1960	0.046068	Asia	8996351.0	414442.773324
53	Afghanistan	AFG	1961	0.053615	Asia	9166764.0	491475.529354
54	Afghanistan	AFG	1962	0.073781	Asia	9345868.0	689550.645811
55	Afghanistan	AFG	1963	0.074251	Asia	9533954.0	707909.126949
56	Afghanistan	AFG	1964	0.086317	Asia	9731361.0	839977.440205

df_hist = pd.DataFrame(df.groupby(by='Country Name', as_index=False)['Whole emission'].mean())
df_hist = df_hist.sort_values(by=['Whole emission'], ascending=False)
df_hist.head()

	Country Name	Whole emission
141	United States	4.699988e+09
28	China	2.639401e+09
74	Japan	9.278755e+08
66	India	7.040976e+08
140	United Kingdom	5.702256e+08

sns.set(style="whitegrid")
plt.figure(figsize=(16, 40))
sns.barplot(x="Whole emission", 
            y="Country Name", 
            data=df_hist)
plt.show()

Which countries have the highest emissions lately ?

# for instance in year 2011
df_lately = df[df.Year == 2011]
df_lately = df_lately.sort_values(by=['Whole emission'], ascending=False)
plt.figure(figsize=(16, 40))
sns.barplot(x='Whole emission', y="Country Name", data=df_lately)
plt.show()

Are the annual emissions decreasing or increasing ?

Let’s select few countries to show the evolution

selected_countries = ['France', 'Israel', 'Switzerland', 'Chile', 'China', 
                      'Colombia', 'United Kingdom', 'United States', 'Brazil', 'Australia']
plt.figure(figsize=(12, 8))
sns.lineplot(x="Year", 
             y='Whole emission', 
             hue="Country Name", 
             data=df[df["Country Name"].isin(selected_countries)])
plt.title('Evolution of the CO2 emissions for few countries')
plt.show()

plt.figure(figsize=(12, 8))
plt.title('Evolution of the emissions for each continent')

sns.lineplot(x="Year", 
             y='Whole emission', 
             hue="Continent", 
             data=df)
plt.show()

df_mean = pd.DataFrame(df.groupby(by=['Continent', 'Year'], as_index=False)['Whole emission'].mean())

plt.figure(figsize=(12, 8))
plt.title('Evolution of the MEAN emissions for each continent')

sns.lineplot(x="Year", 
             y='Whole emission', 
             hue="Continent", 
             data=df_mean)

plt.show()

Evolution of emission share

df_mean_pivot = pd.pivot_table(df_mean, index='Year', values='Whole emission', columns='Continent')
df_mean_pivot.head()

Continent	Africa	Asia	Europe	North America	Oceania	South America
Year
1960	3.536444e+06	3.928689e+07	6.609129e+07	1.219828e+08	1.010739e+07	1.660111e+07
1961	3.695651e+06	3.453619e+07	6.850683e+07	1.217903e+08	1.037739e+07	1.681596e+07
1962	3.804803e+06	3.239489e+07	7.269145e+07	1.265432e+08	1.072653e+07	1.796754e+07
1963	4.057752e+06	3.418465e+07	7.773613e+07	1.316383e+08	1.145539e+07	1.827552e+07
1964	4.494858e+06	3.575509e+07	8.103279e+07	1.384992e+08	1.240671e+07	1.917087e+07

df_mean_perc = df_mean_pivot.divide(df_mean_pivot.sum(axis=1), axis=0)

plt.figure(figsize=(12, 8))

# Make the plot
plt.stackplot(range(1,53),
              df_mean_perc['Africa'], 
              df_mean_perc["Asia"], 
              df_mean_perc["Europe"],
              df_mean_perc["North America"],
              df_mean_perc["Oceania"],
              df_mean_perc["South America"],
              labels=['Africa','Asia','Europe','North America','Oceania','South America'])

# Formatting the plot
plt.legend(loc='upper left')
plt.margins(0,0)
plt.title('Evolution of emissions share over time')
plt.show()

World map