Preparation script to create data for a Sankey diagram in Tableau. Input data is a row based and 'country' and 'sector' are two dimensions of the data.

prep_sankey_tableau.py

# Preparation script to create data for a Sankey diagram in Tableau.
# Input data is a row based and 'country' and 'sector' are two dimensions of the data.
# See https://public.tableau.com/profile/samuelleach#!/vizhome/SankeyTemplate/Dashboard1

import pandas as pd

infile = 'all_loans.csv'
outfile = 'all_loans_sankey.csv'

sankey_columns = ['country', 'sector']

print 'Reading ' + infile
df = pd.read_csv(infile, low_memory=False)

print 'Performing groupby operations'
bygroup_treatment = df.groupby(sankey_columns)

df = bygroup_treatment.sum()

print 'Adding RowType column'
df['RowType'] = pd.Series('Dummy', index=df.index)

print 'Copying dataframe'
df2 = df.copy()
df2['RowType'] = 'Real'

print 'Concatenating dataframes'
result = pd.concat([df, df2])

print 'Writing data frame to ' + outfile
result.to_csv(outfile)