10 Making an effective analysis

 

This chapter covers

  • Planning an analysis
  • Thinking through code, data, and project structure
  • Delivering the analysis to the client

This chapter is written in the context of data scientists who focus on decision science and analytics—people who use data to provide ideas and suggestions to the business. Although machine learning engineers also have to do analysis before building and deploying models, some of the content around stakeholder management with pretty visualizations is less relevant. If you’re a machine learning engineer and you’re reading this book, don’t worry; this chapter is still plenty relevant to you, and you’ll love chapter 11, which covers deploying models into production.

10.1. The request

10.2. The analysis plan

10.3. Doing the analysis

10.3.1. Importing and cleaning data

10.3.2. Data exploration and modeling

10.3.3. Important points for exploring and modeling

10.4. Wrapping it up

10.4.1. Final presentation

10.4.2. Mothballing your work

10.5. Interview with Hilary Parker, data scientist at Stitch Fix

How does thinking about other people help your analysis?

How do you structure your analyses?

What kind of polish do you do in the final version?

How do you handle people asking for adjustments to an analysis?

Summary