Pareto Process

pareto_process.py

# BOX/SKU
df_par = pd.DataFrame(df.groupby(['SKU'])['BOX'].sum())
df_par.columns = ['BOX']

# Sort Values
df_par.sort_values(['BOX'], ascending = False, inplace = True)
df_par.reset_index(inplace = True)

# Cumulative Sum 
df_par['CumSum'] = df_par['BOX'].cumsum()

# % CumSum
df_par['%CumSum'] = (100 * df_par['CumSum']/df_par['BOX'].sum())

# % SKU
df_par['%SKU'] = (100 * (df_par.index + 1).astype(float)/(df_par.index.max() + 1))

print("Pareto Analysis for {:,} unique SKU".format(len(df_par)))
df_par.set_index('SKU').head()