Data preparation exercise - Machine Learning

import pandas as pd

mortgage_testing = pd.read_csv('mortgage-testing-data.csv')

Use the dummy function to convert categorical functions to Boolean columns

dummy = pd.get_dummies(mortgage_testing['derived_sex'])

dummy.head()

concatenate the variables so they all show up as new columns on the dataset

mortgage_testing_new = pd.concat((mortgage_testing, dummy),axis=1)

now we can drop columns that are no longer needed. We are only looking at the male and female columns in this exercise.

mortgage_testing_new = mortgage_testing_new.drop(['Joint'], axis=1)

mortgage_testing_new = mortgage_testing_new.drop(['derived_sex'], axis=1)

mortgage_testing_new = mortgage_testing_new.drop(['Sex Not Available'], axis=1)

mortgage_testing_new

With the output one can notice we are just left with male and female columns

Use the dummy function to convert categorical functions to Boolean columns this time for the conforming loan limit column.

dummy1 = pd.get_dummies(mortgage_testing['conforming_loan_limit'])

dummy1.head()

concatenate the variables so they all show up as new columns on the dataset

mortgage_testing_new1 = pd.concat((mortgage_testing_new, dummy1),axis=1)

now we can drop columns that are no longer needed.

mortgage_testing_new1 = mortgage_testing_new1.drop(['conforming_loan_limit'], axis=1)

mortgage_testing_new1.head()

with the output there are now two new columns conforming =1 and not conforming = NC

Perform the same tasks for the action taken column the category

dummy3 = pd.get_dummies(mortgage_testing['action_taken'])

mortgage_testing_final = pd.concat((mortgage_testing_new1, dummy3),axis=1)

Drop columns we no longer need

mortgage_testing_new1 = mortgage_testing_new1.drop(['action_taken'], axis=1)

mortgage_testing_final.head()

Export to a csv file

mortgage_testing_final.to_csv('mortgage_testing.csv', index=False)