Exploratory Data Analysis.py.html

# coding: utf-8

# In[59]:


#Import Packages

import pandas as pd

import numpy as np

import seaborn as sns

import matplotlib as plt
from matplotlib import pyplot as plt


# In[19]:


df1 = pd.read_csv("/Users/Devan/Desktop/MHE/Datasets/NCHS_-_Death_rates_and_life_expectancy_at_birth.csv")
#I believe this is only based in the US


# In[25]:


#Snapshot of data in dataset

df1.head(10)


# In[32]:


#Delete Blank/NA/NAN Values

df2 = df1.dropna()


# In[34]:


#Snapshot of data in dataset

df2.head(10)


# In[35]:


df2.describe()


# In[42]:


#Create Subset for "All Races" & "Both Sexes"

everyone_df = df2[(df2.Race == "All Races") & (df2.Sex == "Both Sexes")]

everyone_df.head(10)


# In[47]:


#Create Scatterplot to Show Avergae Life Expectancy by Year

x = everyone_df['Average Life Expectancy (Years)']

y = everyone_df['Year']

plt.scatter(x,y, color='b', label = 'Average Life Expectancy')


plt.title('Average Life Expectancy (Years) of Everyone')
plt.xlabel('Years Old')
plt.ylabel('Year')
plt.legend()
plt.show()

#As you can see, life expectancy steadily increases as time progress.
#It appears that life expectancy went down in 1918.  I wonder why this occured?  War, disease, etc.?


# In[45]:


#Create Subset for "Black" & "Both Sexes"

black_df = df2[(df2.Race == "Black") & (df2.Sex == "Both Sexes")]

black_df.head(10)


# In[46]:


#Create Subset for "White" & "Both Sexes"

white_df = df2[(df2.Race == "White") & (df2.Sex == "Both Sexes")]

white_df.head(10)


# In[48]:


#Create Scatterplot to Show Avergae Life Expectancy by Year Comparing Black & White Americans

x = black_df['Average Life Expectancy (Years)']

y = black_df['Year']

plt.scatter(x,y, color='r', label = 'Black Average Life Expectancy')

x2 = white_df['Average Life Expectancy (Years)']
plt.scatter(x2,y, color='b', label = 'White Average Life Expectancy')

plt.title('Life Expectancy by Race')
plt.xlabel('Years Old')
plt.ylabel('Year')
plt.legend()
plt.show()

#According to the graph white americans live longer on average. 
#Is this due to genetics, or is there another root cause (e.g. education, income, etc.)
#Once again 1918 appears as an outlier, this time for white americans. What happened in 1918?
#1918 marked the beginning of the flu pandemic. 28% of americans were infected, and 500,000 to 675,000 died.


# In[52]:


#Add Combined Average Life Expectancy to Scatterplot

x = black_df['Average Life Expectancy (Years)']

y = black_df['Year']

plt.scatter(x,y, color='r', label = 'Black Average Life Expectancy')

x2 = white_df['Average Life Expectancy (Years)']
plt.scatter(x2,y, color='b', label = 'White Average Life Expectancy')

x3 = everyone_df['Average Life Expectancy (Years)']
plt.scatter(x3,y, color='g', label = 'Combined Average Life Expectancy')

plt.title('Life Expectancy by Race')
plt.xlabel('Years Old')
plt.ylabel('Year')
plt.legend()
plt.show()

#It appears that the combined average life expectancy is almost a reflection of the white american life expectancy
#I am going to guess that the white population significantly outnumbered the black populations.


# In[ ]:


#My data is based on years and averages. As such, I don't have individual counts per population (black, white. etc.)
#The next 3 graphs aren't a good pair/match for my data, but I just wanted to show the graphs.
#Let's imagine each year represented a person instead.


# In[68]:


#Create Kernal Density Estimation (KDE) Graph

sns.distplot(black_df['Average Life Expectancy (Years)'], hist=False,color='red',label='black')
sns.distplot(white_df['Average Life Expectancy (Years)'], hist=False, color='blue', label='white')
plt.axvline(linewidth=2, linestyle = '-.', color='red', x=black_df['Average Life Expectancy (Years)'].median())
plt.axvline(linewidth=2, linestyle = '-.', color='blue', x=white_df['Average Life Expectancy (Years)'].median())
plt.legend()
plt.title("Average Life Expectancy KDE Plot", fontsize='20')
plt.xlabel('Average Life Expectancy (Years)')
plt.ylabel('Frequency')
plt.show()


# In[79]:


#Create Violin Plot

sns.violinplot(black_df['Average Life Expectancy (Years)'], color='red', label='black')
sns.violinplot(white_df['Average Life Expectancy (Years)'], color='blue', label='white')
plt.title("Average Life Expectancy \n Violin Plot", fontsize='18') #/n is like hitting "enter" key
plt.axvline(linewidth=1, linestyle = '-.', color='red', x=black_df['Average Life Expectancy (Years)'].median())
plt.axvline(linewidth=1, linestyle = '-.', color='blue', x=white_df['Average Life Expectancy (Years)'].median())
plt.xlabel('Average Life Expectancy (Years)')
plt.show()


# In[76]:


#Create Cumulative Distribution Function Plot (CDF)

sns.kdeplot(black_df['Average Life Expectancy (Years)'], cumulative=True, linewidth=4, color='red',label='black')
sns.kdeplot(white_df['Average Life Expectancy (Years)'], cumulative=True, linewidth=4, color='blue',label='white')
plt.axvline(linewidth=1, linestyle = '-.', color='green', x=65)
plt.title("Average Life Expectancy \n CDF Plot", fontsize='16')
plt.xlabel("Average Life Expectancy (Years)")
plt.ylabel("Cumulative Percent")
plt.ylim(0,1.1) #0=min 1.1=max
plt.show()