Setup

import numpy as np import pandas as pd import matplotlib.pyplot as plt import seaborn as sns from scipy import stats import inequality from inequality.gini import Gini_Spatial import plotly.express as px import plotly.graph_objects as go #chart_studio.tools.set_credentials_file(username='econdata777', api_key='ADDhere') import statsmodels.api as sm import statsmodels.formula.api as smf import warnings #warnings.filterwarnings('ignore')

Import data

df1 = pd.read_csv("https://raw.githubusercontent.com/quarcs-lab/mendez2020-convergence-clubs-code-data/master/assets/dat.csv") #df1 = pd.read_csv("https://raw.githubusercontent.com/quarcs-lab/mendez2020-convergence-clubs-code-data/master/assets/dat.csv", parse_dates =['year'])

df1

idint64

1 - 108

countryobject

Zambia0.9%

Italy0.9%

106 others98.1%

2699

108

Zambia

1199

Italy

1324

Kyrgyz Republic

1299

Kenya

1274

Kazakhstan

2649

106

Vietnam

1249

Japan

1224

Jamaica

1174

Israel

1374

Lithuania

#df1.head()

df1.columns

df1.dtypes

df1['country'].unique()

Dataset definitions

# Import definitions of dataset df1_def = pd.read_csv("https://raw.githubusercontent.com/quarcs-lab/mendez2020-convergence-clubs-code-data/master//assets/dat-definitions.csv") df1_def

var_nameobject

country3.6%

year3.6%

26 others92.9%

var_defobject

Standardized country name (from PWT)3.6%

Year3.6%

26 others92.9%

country

Standardized country name (from PWT)

year

Year

GDP

Physical Capital

pop

Population

Labor Force

Years of Schooling

alpha_it

Variable Capital Share

GDPpc

GDP per capita

Labor Productivity

Descriptive statistics

df1.describe().round(2)

idfloat64

yearfloat64

count

2700

mean

54.5

2002

std

31.18

7.21

min

1990

25%

27.75

1996

50%

54.5

2002

75%

81.25

2008

max

108

2014

select

df1.GDPpc

pivot_table

# Pivot panel data from long form to wide form df1_gdp_wide = df1.pivot_table( index = ['country', 'region', 'hi1990'], columns = 'year', values = 'GDPpc').reset_index(drop=False) # Make sure the column names are strings df1_gdp_wide.columns = df1_gdp_wide.columns.astype(str)

df1_gdp_wide

countryobject

Albania0.9%

Algeria0.9%

106 others98.1%

regionobject

Europe27.8%

Asia25.9%

3 others46.3%

Albania

Europe

Algeria

Africa

Argentina

Americas

Armenia

Asia

Australia

Oceania

Austria

Europe

Bangladesh

Asia

Belgium

Europe

Benin

Africa

Bolivia

Americas

# Pivot panel data from long form to wide form df1_lngdp_wide = df1.pivot_table( index = ['country', 'region', 'hi1990'], columns = 'year', values = 'log_GDPpc').reset_index(drop=False) # Make sure the column names are strings df1_lngdp_wide.columns = df1_lngdp_wide.columns.astype(str)

df1_lngdp_wide['g9014']=df1_lngdp_wide['2014']-df1_lngdp_wide['1990']

df_high = df1_gdp_wide.query("hi1990 == 'yes'") df_low = df1_gdp_wide.query("hi1990 == 'no'") df_highl = df1_lngdp_wide.query("hi1990 == 'yes'") df_lowl = df1_lngdp_wide.query("hi1990 == 'no'")

df1_lngdp_wide

countryobject

Albania0.9%

Algeria0.9%

106 others98.1%

regionobject

Europe27.8%

Asia25.9%

3 others46.3%

Albania

Europe

Algeria

Africa

Argentina

Americas

Armenia

Asia

Australia

Oceania

Austria

Europe

Bangladesh

Asia

Belgium

Europe

Benin

Africa

Bolivia

Americas

df1_lngdp_wide.describe().round(2)

1990float64

1991float64

count

108

mean

8.54

8.56

std

1.19

min

6.16

6.2

25%

7.57

7.59

50%

8.68

8.71

75%

9.6

9.63

max

10.49

10.51

df1_gdp_wide.std().round(2).plot();

<ipython-input-18-816c33b482bc>:1: FutureWarning: Dropping of nuisance columns in DataFrame reductions (with 'numeric_only=None') is deprecated; in a future version this will raise TypeError.  Select only valid columns before calling the reduction.
  df1_gdp_wide.std().round(2).plot();

df_high.std().round(2).plot();

<ipython-input-19-3df1beec0f34>:1: FutureWarning: Dropping of nuisance columns in DataFrame reductions (with 'numeric_only=None') is deprecated; in a future version this will raise TypeError.  Select only valid columns before calling the reduction.
  df_high.std().round(2).plot();

df_low.std().round(2).plot();

<ipython-input-20-91a38291c563>:1: FutureWarning: Dropping of nuisance columns in DataFrame reductions (with 'numeric_only=None') is deprecated; in a future version this will raise TypeError.  Select only valid columns before calling the reduction.
  df_low.std().round(2).plot();

df1_lngdp_wide.loc[:,'1990':'2014'].std().round(2).plot();

Ineuqality

years = np.arange(1990, 2015).astype(str)

Gini

def gini_by_col(column): return inequality.gini.Gini(column.values).g

inequalities = df1_gdp_wide[years].apply(gini_by_col, axis=0).to_frame('gini')

inequalities.plot();

high income

ine = df_high[years].apply(gini_by_col, axis=0).to_frame('gini')

ine.plot();

low income

inel = df_low[years].apply(gini_by_col, axis=0).to_frame('gini')

Theil index

def theil(column): return inequality.theil.Theil(column.values).T

inequalities['theil'] = df1_gdp_wide[years].apply(theil, axis=0).to_frame('theil') ine['theil'] = df_high[years].apply(theil, axis=0).to_frame('theil') inel['theil'] = df_low[years].apply(theil, axis=0).to_frame('theil')

inequalities['theil'].plot();

CV

inequalities['CV'] = df1_gdp_wide[years].apply(stats.variation, axis=0).to_frame('CV') ine['CV'] = df_high[years].apply(stats.variation, axis=0).to_frame('CV') inel['CV'] = df_low[years].apply(stats.variation, axis=0).to_frame('CV')

inequalities['CV'].plot();

df = pd.DataFrame({'year':range(1990, 2015,1), 'gini': inequalities['gini'], 'theil': inequalities['theil'], 'CV': inequalities['CV']}) dfh = pd.DataFrame({'year':range(1990, 2015,1), 'gini': ine['gini'], 'theil': ine['theil'], 'CV': ine['CV']}) dfl = pd.DataFrame({'year':range(1990, 2015,1), 'gini': inel['gini'], 'theil': inel['theil'], 'CV': inel['CV']})

df.set_index('year').plot(figsize=(10,10)); plt.xticks(range(1990, 2015,2))

dfh.set_index('year').plot(figsize=(10,10)); plt.xticks(range(1990, 2015,2))

dfl.set_index('year').plot(figsize=(10,10)); plt.xticks(range(1990, 2015,2))

Visualize data

px.strip(df1, x = 'GDPpc', y = 'region', hover_name= 'country', hover_data= ['h', 'ky'], color= 'region', animation_frame= 'year')

Line plots

px.line(df1, x='year', y='log_GDPpc', color='country', facet_col= 'hi1990', facet_col_wrap= 2, height= 800)

Regression

Overall beta convergence

px.scatter( df1_lngdp_wide, x="1990", y="g9014", color="hi1990", hover_name="country", hover_data= ['region'], trendline="ols", trendline_scope="overall", labels={"1990": "Log GDPpc in 1990", "g9014": "Growth GNIpc 1990-2014"} )

Beta convergence in high income country

px.scatter( df1_lngdp_wide, x="1990", y="g9014", color="hi1990", hover_name="country", trendline="ols", labels={"1990": "Log GDPpc in 1990", "g9014": "Growth GDPpc 1990-2014"})

Beta convergence in low income country

px.scatter( df_highl, x="1990", y="g9014", color="region", hover_name="country", trendline="ols", labels={"1990": "Log GDPpc in 1990", "g9014": "Growth GDPpc 1990-2014"})

px.scatter( df_lowl, x="1990", y="g9014", color="region", hover_name="country", trendline="ols", labels={"1990": "Log GDPpc in 1990", "g9014": "Growth GDPpc 1990-2014"})