Missing data and outliers

Practicing Statistics Interview Questions in Python

Conor Dewey

Data Scientist, Squarespace

Handling missing data

• Drop the whole row
• Impute missing values

Drop the whole row

df.dropna(inplace=True)

 

Impute missing values

• Constant value
• Randomly selected record
• Mean, median, or mode
• Value estimated by another model

A few useful functions

• isnull()
• dropna()
• fillna()

Dealing with outliers

• Standard deviations
• Interquartile range (IQR)

Standard deviations

 

¹ Wikimedia

Interquartile range (IQR)

 

¹ Wikimedia

Summary

• Drop the whole row
• Impute missing values
• Standard deviations
• Interquartile range

Let's prepare for the interview!

Practicing Statistics Interview Questions in Python