!pip install openpyxl

import pandas as pd import numpy as np import plotly.express as px

psid_df = pd.read_csv('psid_task_data.txt', sep='\t') psid_df.head()

cpi_df = pd.read_excel('bls_cpi_data.xlsx') cpi_df.columns = cpi_df.iloc[10] # fix excel parsing errors cpi_df = cpi_df.iloc[11:] cpi_df = cpi_df.set_index('Year') cpi_df.head()

psid_df['birth_year'] = psid_df['year'] - psid_df['age']

basket_2010 = cpi_df.loc[2010]['Annual'] # 2010 annual average adj_cpi_df = cpi_df.divide(basket_2010).multiply(100) adj_cpi_df.head()

stable_psid_df = psid_df[psid_df['change_in_fam_comp']==0].drop('change_in_fam_comp', axis=1) stable_psid_df.head()

adj_psid_df = stable_psid_df.join(adj_cpi_df['Annual'],on='year',how='left').rename({'Annual':'CPI'}, axis=1) adj_psid_df['full_income'] = adj_psid_df['full_income']/adj_psid_df['CPI']*100 # adjust by CPI adj_psid_df.head()

weighted_avg = lambda x: np.average(list(x), weights=list(adj_psid_df.loc[x.index, 'full_weights'])) average_income = adj_psid_df[['year','full_income']].groupby(['year']).aggregate(weighted_avg) average_income.to_csv('average_incomes.csv') average_income.head()

fig = px.line(data_frame=average_income, x=average_income.index, y='full_income', title='Inflation Adjusted Income 1982-2015 for Stable Households', labels=dict(year="Year", full_income='Annual Income in 2010 Dollars')) fig.show()

infl = adj_cpi_df['Annual'].pct_change()[1:] infl.head()

""" For a given year `t`, a birth year `birth`, and lambda `lmbda`, computes the experienced inflation. """ def compute_exp_infl(year, lmbda, birth=min(infl.index)): age = year-birth # consider the current age of the agent def weights(age, lag): # weight experiences in terms of ages return (age-lag)**lmbda/np.sum((np.arange(1, age+1))**lmbda) def wght_inf(lag): # compute weighted inflation for given year lag return weights(age, lag)*infl.loc[year-lag] return sum(map(wght_inf, np.arange(0, age))) # sum weighted inflation experiences

lmbda = 4 # set lambda exp_infl = pd.Series(map(lambda x: compute_exp_infl(x, lmbda), infl.index)) # apply to inflation series exp_infl.index = infl.index # adjust index infl_df = pd.DataFrame({'current':infl, 'experienced':exp_infl}).iloc[1:,].reset_index() # drop first year infl_df = pd.melt(infl_df, id_vars=['Year'], value_vars=['current','experienced']).rename({'variable':'Inflation Type', 'value':'Inflation Level'}, axis=1) px.line(infl_df, x='Year', y='Inflation Level', color='Inflation Type', title=f'Current vs. Experienced Inflation for Lambda={lmbda}')

print("{:.5f}%!".format(compute_exp_infl(year=2017, lmbda=3, birth=1949)*100))