Trans-Atlantic Slave Trade

Download worked project

Browse files online

Two centuries ago the shipping of enslaved Africans across the Atlantic was morally indistinguishable from shipping sugar or textiles. This migration experience covers an era of very dramatic shifts in perceptions of good and evil, which provided the Americas with a crucial labor force for their own economic development.

Data provider: The Trans-Atlantic Slave Trade Database. 2020. SlaveVoyages. https://www.slavevoyages.org. You are enouraged to explore the dataset with the very interesting online tool they built, in particular check out the Maps and Timelapse tabs.

Data license:

• Historical data: The Trans-Atlantic Slave Trade Database. 2020. SlaveVoyages. https://www.slavevoyages.org (accessed June 9, 2020). License: Public domain (use restrictions do not apply).

• Imputed data: Estimates. 2020. SlaveVoyages. https://slavevoyages.org/assessment/estimates (accessed June 9, 2020). License: Creative Commons Attribution-Noncommercial 3.0 United States License.

What to do

1. Unzip exercises zip in a folder, you should obtain something like this:

slave-trade-prj
    slave-trade.ipynb
    slave-trade-sol.ipynb
    slave-trade.csv
    region-codes.csv
    soft.py
    jupman.py

WARNING: to correctly visualize the notebook, it MUST be in an unzipped folder !

2. open Jupyter Notebook from that folder. Two things should open, first a console and then a browser. The browser should show a file list: navigate the list and open the notebook slave-trade.ipynb

3. Go on reading the notebook, and write in the appropriate cells when asked

Shortcut keys:

• to execute Python code inside a Jupyter cell, press Control + Enter

• to execute Python code inside a Jupyter cell AND select next cell, press Shift + Enter

• to execute Python code inside a Jupyter cell AND a create a new cell aftwerwards, press Alt + Enter

• If the notebooks look stuck, try to select Kernel -> Restart

1. read_trade

Each line in slave-trade.csv represents a ship voyage from a purchase place to a landing place. Parse it with a csv reader and output a list of dictionaries, one per voyage according to the output excerpt.

• Each ship has a nation flag NATINIMP

• Each voyage has purchase place code MJBYPTIMP and a landing place code MJSLPTIMP with five digits format xyzvt that indicate a specific town: you MUST save more generic codes of the form xyz00 which indicate broader regions.

• WARNING 1: convert to int only VOYAGEID and YEARAM, leave MJBYPTIMP and MJSLPTIMP as strings

• WARNING 2: some codes in slave-trade.csv have a space instead of a number, in those cases save code 00000

import pandas as pd
import numpy as np
df = pd.read_csv('slave-trade.csv', encoding='UTF-8')
df[df.VOYAGEID.isin([1, 2024, 2393, 4190])]

	VOYAGEID	YEARAM	NATINIMP	MJBYPTIMP	MJSLPTIMP
0	1	1817	Portugal/Brazil	60820	50299
2000	2024	1840	U.S.A.	60615	31399
2361	2393	1829	Spain/Uruguay	60212
4000	4190	1854	U.S.A.	60515	31301

Region labels: For each location you need to also save its label, which you can find in separate file region-codes.csv (load the file with a csv reader)

• WARNING 1: in region-codes.csv there are only codes in format xyz00

• WARNING 2: some region codes are missing, in those cases place label 'unknown'

import pandas as pd
dfr = pd.read_csv('region-codes.csv', encoding='UTF-8', dtype=str)
dfr[dfr.Value.isin(['60800','60600','31300','50200','60500'])]

	Value	Region
47	31300	Cuba
84	50200	Bahia
92	60500	Bight of Benin
93	60600	Bight of Biafra and Gulf of Guinea islands
95	60800	Southeast Africa and Indian Ocean islands

Output excerpt: (for full output see expected_db.py)

>>> read_voyages('slave-trade.csv', 'region-codes.csv')
[
{'flag': 'Portugal/Brazil',
 'id': 1,
 'landing_id': '50200',
 'landing_label': 'Bahia',
 'purchase_id': '60800',
 'purchase_label': 'Southeast Africa and Indian Ocean islands',
 'year': 1817},
{'flag': 'U.S.A.',
 'id': 2024,
 'landing_id': '31300',
 'landing_label': 'Cuba',
 'purchase_id': '60600',
 'purchase_label': 'Bight of Biafra and Gulf of Guinea islands',
 'year': 1840},
{'flag': 'Spain/Uruguay',
 'id': 2393,
 'landing_id': '00000',
 'landing_label': 'unknown',
 'purchase_id': '60200',
 'purchase_label': 'Sierra Leone',
 'year': 1829},
{'flag': 'U.S.A.',
 'id': 4190,
 'landing_id': '31300',
 'landing_label': 'Cuba',
 'purchase_id': '60500',
 'purchase_label': 'Bight of Benin',
 'year': 1854},
  .
  .
]

Show solution

import csv

def read_voyages(slave_trade_csv, region_codes_csv):


    with open(region_codes_csv, encoding='utf-8', newline='') as fregions:
        my_reader = csv.DictReader(fregions, delimiter=',')

        regions_db = {}
        for d in my_reader:
            regions_db[d['Value']] = d['Region']

    with open(slave_trade_csv, encoding='utf-8', newline='') as f:
        my_reader = csv.DictReader(f, delimiter=',')

        ret = []

        for d in my_reader:

            voyage = {}
            voyage['id'] = int(d['VOYAGEID'])
            voyage['year'] = int(d['YEARAM'])
            voyage['flag'] = d['NATINIMP']

            if  d['MJBYPTIMP'].strip():
                pur_reg = d['MJBYPTIMP'][:3] + '00'
                voyage['purchase_id'] = pur_reg
                if pur_reg in regions_db:
                    voyage['purchase_label'] = regions_db[pur_reg]
                else:
                    voyage['purchase_label'] = 'unknown'
            else:
                voyage['purchase_id'] = '00000'
                voyage['purchase_label'] = 'unknown'

            if d['MJSLPTIMP'].strip():
                lan_reg = d['MJSLPTIMP'][:3] + '00'
                voyage['landing_id'] = lan_reg
                if lan_reg in regions_db:
                    voyage['landing_label'] = regions_db[lan_reg]
                else:
                    voyage['landing_label'] = 'unknown'
            else:
                voyage['landing_id'] = '00000'
                voyage['landing_label'] = 'unknown'

            ret.append(voyage)

        return ret


voyages_db = read_voyages('slave-trade.csv', 'region-codes.csv')

print('OUTPUT EXCERPT:')
from pprint import pformat
print('[\n' +',\n'.join([pformat(voyages_db[vid]) for vid in [0,2000,2361, 4000]]) + ',\n  .\n  .\n]')

OUTPUT EXCERPT:
[
{'flag': 'Portugal/Brazil',
 'id': 1,
 'landing_id': '50200',
 'landing_label': 'Bahia',
 'purchase_id': '60800',
 'purchase_label': 'Southeast Africa and Indian Ocean islands',
 'year': 1817},
{'flag': 'U.S.A.',
 'id': 2024,
 'landing_id': '31300',
 'landing_label': 'Cuba',
 'purchase_id': '60600',
 'purchase_label': 'Bight of Biafra and Gulf of Guinea islands',
 'year': 1840},
{'flag': 'Spain/Uruguay',
 'id': 2393,
 'landing_id': '00000',
 'landing_label': 'unknown',
 'purchase_id': '60200',
 'purchase_label': 'Sierra Leone',
 'year': 1829},
{'flag': 'U.S.A.',
 'id': 4190,
 'landing_id': '31300',
 'landing_label': 'Cuba',
 'purchase_id': '60500',
 'purchase_label': 'Bight of Benin',
 'year': 1854},
  .
  .
]

import csv

def read_voyages(slave_trade_csv, region_codes_csv):
    raise Exception('TODO IMPLEMENT ME !')

voyages_db = read_voyages('slave-trade.csv', 'region-codes.csv')

print('OUTPUT EXCERPT:')
from pprint import pformat
print('[\n' +',\n'.join([pformat(voyages_db[vid]) for vid in [0,2000,2361, 4000]]) + ',\n  .\n  .\n]')

# TESTING
from pprint import pformat; from expected_db import expected_db
for i in range(0, len(expected_db)):
    if expected_db[i] != voyages_db[i]:
        print('\nERROR at index', i, ':')
        print('  ACTUAL:\n', pformat(voyages_db[i]))
        print('  EXPECTED:\n', pformat(expected_db[i]))
        break
if len(voyages_db) != len(expected_db):
    print("ERROR: Different lengths!  voyages_db:", len(voyages_db), "   expected_db:", len(expected_db))

2. Deportation

For each link purchase -> landing place, count in how many voyages it was present, then draw result in networkx.

• as edge weight use a normalized value from 0.0 to 1.0 (maximal count found in the graph)

• show only edges with weight greater or equal to min_weight

• to display the graph from right to left, set G.graph['graph']= {'rankdir':'RL'}

• for networkx attributes see this example, make sure to display edges proportionally to the weight

Example:

>>> show_deportation(voyages_db, 0.09)
COUNTS EXCERPT SOLUTION:
{
  ('60800', '50200') : 48,
  ('60700', '50200') : 1301,
  ('60700', '50400') : 2770,
  ('60800', '50400') : 443,
  ('60900', '50400') : 196,
    .
    .
}

Show solution

import networkx as nx
from soft import draw_nx

def show_deportation(voyages, min_weight):


    edges = {}

    for v in voyages_db:
        t = (v['purchase_id'], v['landing_id'])
        if t in edges:
            edges[t] += 1
        else:
            edges[t] = 1



    G = nx.DiGraph()
    G.graph['graph']= {'rankdir':'RL'}    # force horiz right to left layout

    mx = max(edges.values())
    for v in voyages_db:
        s = edges[(v['purchase_id'], v['landing_id'])]
        r = s / mx
        if r >= min_weight:
            G.add_node(v['purchase_id'], fontcolor='black', color='brown', label=v['purchase_label'])
            G.add_node(v['landing_id'], fontcolor='darkorange', color='orange', label=v['landing_label'])
            G.add_edge(v['purchase_id'],v['landing_id'],
                       color="blue",
                       penwidth= 7 * r,
                       weight=r)

    draw_nx(G,

           )

show_deportation(voyages_db, 0.09)
#show_deportation(voyages_db, 0.06)

COUNTS EXCERPT SOLUTION:
{
  ('60800', '50200') : 48,
  ('60700', '50200') : 1301,
  ('60700', '50400') : 2770,
  ('60800', '50400') : 443,
  ('60900', '50400') : 196,
    .
    .
}
Image saved to file:  expected-deportation-plot.png

import networkx as nx
from soft import draw_nx

def show_deportation(voyages, min_weight):
    raise Exception('TODO IMPLEMENT ME !')
show_deportation(voyages_db, 0.09)
#show_deportation(voyages_db, 0.06)

3. The time to stop

Given a nation flag, plot inside draw_time the number of voyages per year done by ships belonging to that flag.

DO NOT call plt.show nor plt.figure

• we show some counts example but to calculate the data feel free to use any method you want

• to associate a plot to a label, use i.e. plt.plot(xs, ys, label='France')

Example:

>>> fig = plt.figure(figsize=(15,6))
>>> draw_time(voyages_db, 'France')
>>> draw_time(voyages_db, 'U.S.A.')
>>> draw_time(voyages_db, 'Great Britain')
>>> plt.legend()
>>> plt.show()
France COUNTS EXCERPT SOLUTION:
{
        1816:7,
       1819:30,
       1821:59,
       .
       .
              }
U.S.A. COUNTS EXCERPT SOLUTION:
{
        1821:1,
        1827:1,
        1837:2,
       .
       .
              }
Great Britain COUNTS EXCERPT SOLUTION:
{
        1810:1,
        1809:1,
        1811:1,
       .
       .
              }

Show solution

%matplotlib inline
import matplotlib.pyplot as plt

def draw_time(voyages, flag):


    plt.title("Slave trade voyages SOLUTION")
    plt.xlabel('years')
    plt.ylabel('voyages')

    counts = {}
    for v in voyages:
        if v['flag'] == flag:
            y = v['year']
            if y in counts:
                counts[y] += 1
            else:
                counts[y] = 1



    xs = sorted(counts.keys())
    ys = [counts[x] for x in xs]
    plt.plot(xs, ys, label=flag)


fig = plt.figure(figsize=(15,6))
draw_time(voyages_db, 'France')
draw_time(voyages_db, 'U.S.A.')
draw_time(voyages_db, 'Great Britain')
plt.legend()


plt.show()

France COUNTS EXCERPT SOLUTION:
{
        1816:7,
       1819:30,
       1821:59,
       .
       .
              }
U.S.A. COUNTS EXCERPT SOLUTION:
{
        1821:1,
        1827:1,
        1837:2,
       .
       .
              }
Great Britain COUNTS EXCERPT SOLUTION:
{
        1810:1,
        1809:1,
        1811:1,
       .
       .
              }

%matplotlib inline
import matplotlib.pyplot as plt

def draw_time(voyages, flag):
    raise Exception('TODO IMPLEMENT ME !')

fig = plt.figure(figsize=(15,6))
draw_time(voyages_db, 'France')
draw_time(voyages_db, 'U.S.A.')
draw_time(voyages_db, 'Great Britain')
plt.legend()


plt.show()