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

# # Making a choropleth

# Choropleths are a way to represent some non-geographical data that varies by regions on your map. We'll be making choropleths with Folium and Geopandas. 
# 
# We're going to start with a dataset from the World Bank on Internet usage in 2016.

# In[1]:


import pandas
import matplotlib.pyplot as plt


# In[2]:


internet = pandas.read_csv('data/world_internet_usage_2016.csv', skiprows=18, names=["name", "internet_use"])
internet.head()


# In[3]:


internet[internet.internet_use.isnull()]


# In[4]:


internet.dropna(inplace=True)
internet.head()


# ### Geopandas

# In[5]:


import geopandas
world = geopandas.read_file(geopandas.datasets.get_path('naturalearth_lowres'))
world.head()


# In[6]:


len(set(world.name) & set(internet.name))


# In[7]:


len(set(world.name) - set(internet.name))


# In[8]:


len(set(internet.name) - set(world.name))


# In[9]:


world = world.merge(internet, how='left', on='name')
world.head()


# I'm going to copy over some Proj4 strings, that we can use to tell the underlying projection libraries, PyProj and Proj.4, how to project our map.

# In[10]:


mercator = '+proj=merc'
robinson = '+proj=robin'
wgs84 = '+proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs'


# Next, we'll clean up the map a bit by removing Antarctica (makes projecting a bit more difficult, and is irrelevant to what we're making anyway) and any filling any missing data about Internet connectivity with a zero.

# In[11]:


world = world[(world.name!="Antarctica")].fillna(0)


# We'll make a copy of the DataFrame when we do the projections, so we don't get any weird geometric artifacts.

# In[12]:


# mercator projection
worldMercator = world.to_crs(mercator)
worldMercator.plot(figsize=(20,20), column='internet_use', cmap="OrRd", scheme="QUANTILES", k=5, legend=True)
plt.axis('off')
plt.title("Internet usage by country")
plt.savefig('data/internet_choropleth_geopandas.png')
plt.show()


# ### Data munging
# 
# ...

# ### PySAL
# 
# ...

# ### Folium
# 
# ...

# ### References
# 
# Folium choropleth: http://python-visualization.github.io/folium/quickstart.html#Choropleth-maps
# 
# World bank data: https://data.worldbank.org/data-catalog/country-profiles