World Population Trend Analysis using Python Pandas and Seaborn Charts

Download population data from data.gov

populationbycountry19802010millions.csv

Contents

Ø  Read data

Ø  Cleanup data

Ø  World Population as of 2010 – Bar plot

Ø  Population Trend Analysis.

o   Point Plot

o   Box Plot

o   Bar Plot

Ø  Population Growth Rate and Percentage.

o   Growth Rate - Bar Plot

o   Growth Percent - Bar Plot

Read data.

 ## Read data  
 ##  
 import pandas as pd  
 pop=pd.read_csv("populationbycountry19802010millions.csv")  

Cleanup data.

Fill na, rename and convert data type

 pop=pop.fillna(0)  
 pop=pop.replace('--',0)  
 pop=pop.rename(columns={"Unnamed: 0": "geo_location"})  
 ##  
 for i in range(2008,2011):  
   pop[str(i)]=pop[str(i)].astype('float')  

Select top 20 geo locations based on 2010 population

 pop=pop.nlargest(10, '2010').reset_index(drop=True)  
 pop.head()  

World Population Comparison - Year 2010.

 pop=pop.nlargest(10, '2010').reset_index(drop=True)  
 %matplotlib inline  
 import seaborn as sns  
 import matplotlib.pyplot as plt  
 plt.figure( figsize = (20,6))  
 #set style of plots  
 sns.set_style('white')  
 sns.plt.xlabel('geo_location',weight='bold',size=22)  
 sns.plt.xticks(weight='bold',size=15,rotation=-45,color='blue')  
 sns.plt.yticks(weight='bold',size=15,color='blue')  
 #####################################  
 g=sns.barplot( y = '2010',  
      x = 'geo_location',  
      data = pop)  
 ##  
 g.set_title('World Population 2010',size=40)  
 g.set_ylabel('Population in Millions',size=22,weight='bold')  
 g.set_xlabel('Geographic Region',size=22,weight='bold')  

World Population Trend Analysis - Point Plot

 v=pop.columns.drop('geo_location')  
 popM=pd.melt(pop, id_vars='geo_location', value_vars=v)  
 popM['value']=popM['value'].astype('float')  
 #popM=popM[popM['geo_location']in('World')]  
 popM=popM.set_index('geo_location')  
 popM=popM.loc[['World','Asia & Oceania','Africa','China','India','United States']].reset_index()  
 popM.head()  

 plt.figure( figsize = (20,8))  
 #set style of plots  
 sns.set_style('white')  
 sns.plt.xlabel('geo_location',weight='bold',size=22)  
 sns.plt.xticks(weight='bold',size=15,rotation=-90,color='blue')  
 sns.plt.yticks(weight='bold',size=15,color='blue')  
 #####################################  
 g=sns.pointplot(x='variable', y='value', hue='geo_location',  
      data = popM)  
 ##  
 g.set_title('Population Growth Trend (1980 - 2010)',size=30)  
 g.set_ylabel('Population in Millions',size=22,weight='bold')  
 g.set_xlabel('Year',size=22,weight='bold')  
 ##  
 plt.setp(g.get_legend().get_texts(), fontsize='15') # for legend text  
 plt.setp(g.get_legend().get_title(), fontsize='18')  

World Population Trend Analysis - Box Plot

 plt.figure( figsize = (20,10))  
 #set style of plots  
 sns.set_style('white')  
 sns.plt.xlabel('geo_location',weight='bold',size=22)  
 sns.plt.xticks(weight='bold',size=20,rotation=-90,color='blue')  
 sns.plt.yticks(weight='bold',size=15,color='blue')  
 #####################################  
 g=sns.boxplot( x = popM['geo_location'],y = popM['value'])  
 ##  
 g.set_title('Population Growth Trend (1980 - 2010)',size=30)  
 g.set_ylabel('Population in Millions',size=22,weight='bold')  
 g.set_xlabel('Geographic Region',size=22,weight='bold')  
 ##  

World Population Trend Analysis - Bar Plot

 plt.figure( figsize = (20,10))  
 ##set style of plots  
 sns.set_style('white')  
 sns.plt.xlabel('geo_location',weight='bold',size=22)  
 sns.plt.xticks(weight='bold',size=20,rotation=-90,color='blue')  
 sns.plt.yticks(weight='bold',size=15,color='blue')  
 #####################################  
 g=sns.barplot(x='geo_location', y='value', hue='variable', data=popM);  
 ##  
 g.set_title('Population Growth Trend (1980 - 2010)',size=30)  
 g.set_ylabel('Population in Millions',size=22,weight='bold')  
 g.set_xlabel('Geographic Region',size=22,weight='bold')  
 ##  

World Population Growth Rate and Percent

 ##### Prepare Growth Rate and Percentage over the years  
 import numpy as np  
 popM.head()  
 popM['variable']=popM['variable'].astype('int')  
 popM_prev=popM.copy()  
 popM_prev=popM_prev.rename(columns={"value": "value_prev"})  
 popM_prev['variable']=popM_prev['variable'] +1  
 popM=popM.merge(popM_prev,on=['geo_location','variable'],how='left')  
 ##  
 popM=popM.fillna(0)  
 popM['growth']=np.where(popM['value_prev']==0,0,popM['value']-popM['value_prev'])  
 popM['growth_percent']=(popM['growth']/popM['value_prev'])*100  
 popM=popM[['geo_location','variable','value','growth','growth_percent']]  
 popD=popM[popM['variable']!=1980]  
 popD.head()  

World Population Growth Rate - Bar Plot

 plt.figure( figsize = (20,10))  
 ##set style of plots  
 sns.set_style('white')  
 sns.plt.xlabel('geo_location',weight='bold',size=22)  
 sns.plt.xticks(weight='bold',size=20,rotation=-90,color='blue')  
 sns.plt.yticks(weight='bold',size=15,color='blue')  
 #####################################  
 g=sns.barplot(x='geo_location', y='growth', hue='variable', data=popD);  
 ##  
 g.set_title('Population Growth Rate (1980 - 2010)',size=30)  
 g.set_ylabel('Population in Millions',size=22,weight='bold')  
 g.set_xlabel('Geographic Region',size=22,weight='bold')  
 ##  

World Population Growth Percent - Bar Plot

 ##set style of plots  
 plt.figure( figsize = (20,15))  
 sns.set_style('white')  
 sns.plt.xlabel('geo_location',weight='bold',size=22)  
 sns.plt.xticks(weight='bold',size=20,rotation=-90,color='blue')  
 sns.plt.yticks(weight='bold',size=15,color='blue')  
 #####################################  
 g=sns.barplot(x='geo_location', y='growth_percent', hue='variable', data=popD);  
 ##  
 g.set_title('Population Growth Rate Percent (1980 - 2010)',size=30)  
 g.set_ylabel('Population Rate Percent',size=22,weight='bold')  
 g.set_xlabel('Geographic Region',size=22,weight='bold')  
 ##