first use of datachile API

datachile_first.py

import pandas as pd
import matplotlib.pyplot as plt
import squarify    # pip install squarify (algorithm for treemap)
from datachile import ChileCube

client = ChileCube()
query = client.get("exports",
    {"drilldowns": [
            ["Date", "Year"],
            ["Destination Country", "Country", "Country"]
        ],
        "measures": ["FOB US"],
        "cuts": [{"drilldown": ["Date", "Year"],"values": [2016]}],
        "parents": True
    })
###############################
df = pd.DataFrame(query['data'])
df = df[df['FOB US']>df['FOB US'].quantile(.85)]   # remove smaller countries
df = df.sort_values('FOB US', ascending=False)		
df['perc'] = df['FOB US']/df['FOB US'].sum()
df['cp'] = ['%s (%.2f%%)' %(c,100*p) for c,p in zip(df['Country'],df['perc'])]
squarify.plot(sizes=df['FOB US'], label=df['cp'], alpha=.8 )
plt.axis('off')
plt.title('Principales Exportaciones desde Chile año 2016')
plt.show()

My first example of using the datachile API. It's fugly for sure, but we're getting somewhere. Somehow surprised to see that bokeh, seaborn and dash do not provide treemaps as a standard feature.