Question

In Python (using pandas and numpy) I am trying to clean CSV data so it adheres to a strict coding system instead of free response. More specifically, how would I code a simple rule based system to handle the various spellings and word choices that represent the following statuses:

• Never married
• Divorced
• Married
• Widowed
• Separated

Answer 1

#Here is how I converted a string pandas dataframe column to strictly adhere to the categories I have 

import pandas as pd    #Import pandas library
import numpy as np     #Import numpylibrary

#Sample string dataframe
df_sample = pd.DataFrame(pd.Series(['Never married', 'Separated', 'Widowed','Divorced','Married','kjskd','Married','Never married','uguyd']))
print(df_sample)
print(type(df_sample[0]))

#Converting the string dataframe column to a categorical while structurally ordering it using pd.Categorical method
ordered_marstat = ['Never married', 'Divorced', 'Married', 'Widowed', 'Separated']
df = pd.DataFrame(pd.Categorical(df_sample[0], ordered=True,categories=ordered_marstat))
print(df)

This was from my original project where I had to store the string data frame into a new dataframe but you can also replace it with existing column using below syntax

df_sample[0] = pd.DataFrame(pd.Categorical(df_sample[0], ordered=True,categories=ordered_marstat))