#!/usr/bin/env python
# coding: utf-8

# # Global Carbon Budget: Fossil Fuel and Cement Production

# In[1]:


get_ipython().run_line_magic('matplotlib', 'inline')


# In[2]:


import math

import matplotlib.pyplot as plt
plt.rcParams.update({'mathtext.default': 'regular'})
plt.style.use("ggplot")

from pandas_datapackage_reader import read_datapackage


# Let's load the global fossil fuel and cement production emissions data from the [Global Carbon Budget](http://www.globalcarbonproject.org/carbonbudget/index.htm). It has been pre-processed for easy reading in as a [data package](https://github.com/openclimatedata/global-carbon-budget).

# In[3]:


ffc = read_datapackage("https://github.com/openclimatedata/global-carbon-budget",
                       "fossil-fuel-cement")


# In[4]:


ffc.head()


# Turn it into a wide table, with one variable per column.

# In[5]:


ffc = ffc.drop("Source", axis=1).reset_index().pivot(columns="Category", values="Value", index="Year")


# Conversion to Gt and to tonnes CO$_2$ instead of tonnes C.

# In[6]:


ffc_Gt_CO2 = ffc[["Total", "Coal", "Oil", "Gas", "Cement", "Flaring"]] * 3.664 / 1000


# In[7]:


ffc_Gt_CO2[["Coal", "Oil", "Gas", "Cement", "Flaring"]].plot.area(stacked=True)
plt.title("Global Emissions from Fossil Fuel and Cement Production (Data: Global Carbon Budget)")
plt.ylabel("Gt CO$_2$");


# In[8]:


subcategories = ffc_Gt_CO2[["Coal", "Oil", "Gas", "Cement", "Flaring"]] 
ymax = math.ceil(subcategories.max(axis=1).max())
axes = subcategories.plot.area(subplots=True, layout=(3,2), ylim=(0, ymax))
plt.suptitle("Global Emissions from Fossil Fuel and Cement Production in Gt CO$_2$ (Data: Global Carbon Budget)");


# ## Per-capita emissions

# Per-Capita emissions are included in the dataset. We convert to t CO$_2$ per person per year.

# In[9]:


per_capita_CO2 = read_datapackage("https://github.com/openclimatedata/global-carbon-budget",
                       "fossil-fuel-cement-per-capita").drop("Source", axis=1) * 3.664


# In[10]:


per_capita_CO2.plot()
plt.title("Per-Capita Emissions in t CO$_2$ per person per year (Source: Global Carbon Budget)");


# Let's pull in a [population data package](https://github.com/datasets/population) to have a look at the population growth.

# In[11]:


population = read_datapackage("https://github.com/datasets/population")


# In[12]:


world_population = population[population["Country Name"] == "World"][["Year", "Value"]].set_index("Year")
world_population = world_population.Value
(world_population / 10**9).plot()
plt.legend("")
plt.title("World population in billion (Source: World Bank)");


# In[13]:


ffc_t_CO2 = ffc_Gt_CO2 * 10**9


# Re-creating the above plot of per-capita emissions by doing the calculation with the new dataset and plotting them together shows minor differences due to using different population data.

# In[14]:


ax = ffc_t_CO2.Total.loc[1960:].divide(world_population).plot()
per_capita_CO2.plot(ax=ax)
plt.legend(["Global Carbon Budget", "Global Carbon Budget/World Bank"],
          loc="upper left")
plt.title("Per-Capita Emissions in t CO$_2$ per person per year (Source: Global Carbon Budget/World Bank)");


# Let's look at what the per-capita emissions consist of:

# In[15]:


per_capita_categories = ffc_t_CO2[["Coal", "Oil", "Gas", "Cement"]].loc[1960:].divide(
    world_population, axis=0)

per_capita_categories.plot(subplots=True, layout=(2,2), ylim=(0, 2.3))

plt.suptitle("Per-Capita Emissions in t CO$_2$ per person per year (Source: Global Carbon Budget/World Bank)");


# In[ ]: