Your Back to School Todo list

there,

It's been a while!

Like many in the northern hemisphere, I took a summer break. Now I'm back with hundreds of new ideas, and full of energy for the new dataviz season.

This time of year makes me think of everyone starting a new data analyst role. Or deciding to level up their skills after the holidays.

Twelve years ago I was there, beginning my career as a data analyst in academia, studying the wheat genome. If I had to start again, there are so many things I would do differently!

So let me share a few things I wish someone had told me to learn back then. You can use this as a to-do list!

1​️⃣ Write cleaner code

You don't need to be a hardcore programmer or read a 500+ page manual.

Just a few simple habits can be game changers:

• Use the tidyverse → smoother learning curve (R users, Python, check polars)
• Write functions → stop copy-pasting the same code
• Install a code formatter → stop wasting time on spacing
• Master your IDE → shortcuts save seconds, thousands of times
• Split work into manageable pieces with source(), .rds and so on

2​️⃣ Follow reproducibility best practices

• Organize your file structure
• Keep projects self-contained
• Never modify raw data
• Avoid absolute paths (use R Projects instead if using RStudio)
• Manage packages carefully (versions, proper call)

3​️⃣ Generate reports

Stop leaving behind messy R or Python scripts.

Use Quarto to create clean reports that mix code, explanations, results, and visuals. All in one place!!

If you don't know Quarto yet, this will be a game changer. I wrote the shortest intro, and I swear it will take less than 3 minutes to try.

Try Quarto in 3 minutes

And if you know Quarto already, did you know you can build crazy webpages with it?

Note: this works for Python users too. My guess is that Quarto or Marimo will replace Jupyter in the long run!

4​️⃣ Learn Git & GitHub

GitHub is a website that allows you to store your code online.

It also tracks all your changes, allowing you to go back in time. So you won't have to save some ugly results_2_final_final.png files just in case anymore.

It's also the best way, used by millions of developers, to share and collaborate on code.

Last but not least:

• if you're looking for a job, you MUST have a good GitHub profile
• GitHub transforms your work into a little stunning website! (examples)

Done!

The truth is: you've probably heard of these tools. But using them well is what makes you faster, clearer, and more confident as an analyst.

This topic is so important to me and I've wasted so many hours of my life ignoring them.

So I decided to write a structured, efficient resource to summarise everything: productive-r-workflow.com.

See you next week with a short, actionable dataviz tip!
Yan