Demographic data analizer

Created by Darío López Díaz. Work in progress..

The idea of this project is to analize demographic data, which consist of education, race, income and work time per week data of people from different countries. We will compute several values of interest, along the data. The data frame is coming from: Dua, D. and Graff, C. (2019). UCI Machine Learning Repository [http://archive.ics.uci.edu/ml]. Irvine, CA: University of California, School of Information and Computer Science.

import pandas as pd import numpy as np import seaborn as sns import matplotlib.pyplot as plt

The data looks in the following way:

df = pd.read_csv('adult.data.csv') df.head()

print(df.groupby('race').size()) sns.countplot(data=df,x='race') plt.xticks(rotation=45) plt.show()

race
Amer-Indian-Eskimo      311
Asian-Pac-Islander     1039
Black                  3124
Other                   271
White                 27816
dtype: int64

df[['age','sex']].set_index('sex').drop(index='Female').mean()

df.filter(items=['education']).value_counts()['Bachelors'] / df.filter(items=['education']).value_counts().sum() * 100

Salary_Degree = df.filter(items=['education','salary']).value_counts()['Bachelors']+df.filter(items=['education','salary']).value_counts()['Masters']+df.filter(items=['education','salary']).value_counts()['Doctorate'] NumOf_Degree = df.filter(items=['education']).value_counts()['Bachelors']+df.filter(items=['education']).value_counts()['Masters']+df.filter(items=['education']).value_counts()['Doctorate'] ((Salary_Degree / NumOf_Degree)*100)['>50K']

Non_Advance = df.filter(items=['education','salary']).value_counts().drop(labels=['Bachelors','Masters','Doctorate']) Non_Advance_More50 = Non_Advance.drop(labels=['<=50K'],level=1).sum() Non_Advance_Less50 = Non_Advance.drop(labels=['>50K'],level=1).sum() Non_Advance_More50 / (Non_Advance_More50 + Non_Advance_Less50) * 100

/shared-libs/python3.7/py/lib/python3.7/site-packages/pandas/core/generic.py:4153: PerformanceWarning: dropping on a non-lexsorted multi-index without a level parameter may impact performance.
  obj = obj._drop_axis(labels, axis, level=level, errors=errors)

df['hours-per-week'].min()

Min_Hours_More50 = df[df['hours-per-week']==1]['salary'].value_counts()['>50K'] Min_Hours_Total = df[df['hours-per-week']==1]['salary'].value_counts().sum() Min_Hours_More50 / Min_Hours_Total *100

People_Over50byCountry = df[['native-country','salary']].set_index('salary').drop(labels='<=50K').groupby('native-country').size() print(People_Over50byCountry.idxmax()) People_Over50byCountry.max() / People_Over50byCountry.sum() * 100

United-States

People_More50 = df[['occupation','native-country','salary']].set_index('salary').drop(labels='<=50K') People_More50_India = People_More50.set_index('native-country').loc['India'] People_More50_India.groupby('occupation').size().idxmax()