Global Slavery Index Ranking

This notebook shows how to retrieve the country rankings from the Global Slavery Index and save it as a CSV file.

First import the required libraries.

In [1]:

from lxml import html
import pandas as pd
import geonamescache

gc = geonamescache.GeonamesCache()

Find the 5th HTML table on the page and convert it into a Pandas data frame. Print the 1st few rows to inspect the data.

In [2]:

url ='http://www.globalslaveryindex.org/findings/'
xpath = '//table'

tree = html.parse(url)
table = tree.xpath(xpath)[4]
raw_html = html.tostring(table)

df = pd.read_html(raw_html)[0]
df.head(10)

Out[2]:

	Rank	Country Name	Population	Calculated Number of Enslaved	Estimated Enslaved (Lower Range)	Estimate Enslaved (Upper Range)
0	1	Mauritania	3796141	151353	140000	160000
1	2	Haiti	10173775	209165	200000	220000
2	3	Pakistan	179160111	2127132	2000000	2200000
3	4	India	1236686732	13956010	13300000	14700000
4	5	Nepal	27474377	258806	250000	270000
5	6	Moldova	3559541	33325	32000	35000
6	7	Benin	10050702	80371	76000	84000
7	8	Côte d'Ivoire	19839750	156827	150000	160000
8	9	Gambia	1791225	14046	13000	15000
9	10	Gabon	1632572	13707	13000	14000

10 rows × 6 columns

Add an iso3 column containing the 3-letter country codes. Map country names from the Global Slavery Web site that differ from those used in the GeoNames DB with the corresponding names beforehand. Then calculate the percentage of slaves relative to the total population for all countries in the dataset.

In [3]:

cnames = gc.get_countries_by_names()
mapping = {
    u"Côte d'Ivoire": 'Ivory Coast',
    'Republic of Congo': 'Democratic Republic of the Congo',
    'Timor-Leste': 'East Timor',
    'Hong Kong, SAR China': 'Hong Kong'
}

def get_iso3(name):
    if name in mapping:
        name = mapping[name]
    return cnames[name]['iso3']

df['iso3'] = df['Country Name'].apply(get_iso3)
df['Calculated Percentage of Enslaved'] = df['Calculated Number of Enslaved'] / df['Population'] * 100

The modified data frame now has 2 additional columns needed to create the interactive map with D3.

In [4]:

df.head(10)

Out[4]:

	Rank	Country Name	Population	Calculated Number of Enslaved	Estimated Enslaved (Lower Range)	Estimate Enslaved (Upper Range)	iso3	Calculated Percentage of Enslaved
0	1	Mauritania	3796141	151353	140000	160000	MRT	3.987023
1	2	Haiti	10173775	209165	200000	220000	HTI	2.055923
2	3	Pakistan	179160111	2127132	2000000	2200000	PAK	1.187280
3	4	India	1236686732	13956010	13300000	14700000	IND	1.128500
4	5	Nepal	27474377	258806	250000	270000	NPL	0.941990
5	6	Moldova	3559541	33325	32000	35000	MDA	0.936216
6	7	Benin	10050702	80371	76000	84000	BEN	0.799656
7	8	Côte d'Ivoire	19839750	156827	150000	160000	CIV	0.790469
8	9	Gambia	1791225	14046	13000	15000	GMB	0.784156
9	10	Gabon	1632572	13707	13000	14000	GAB	0.839595

10 rows × 8 columns

Save the full data frame and one containing only the columns to display on the gobal slavery map as CSV files. Make sure to set the correct encoding.

In [5]:

df.to_csv('../static/data/csv/globalslaveryindex-full.csv', encoding='utf-8')
df[['iso3', 'Calculated Number of Enslaved', 'Calculated Percentage of Enslaved']].to_csv('../static/data/csv/globalslaveryindex.csv', encoding='utf-8')