What is exploratory data analysis, and what do I do first with a new table?
Meet the seven questions of exploratory analysis you'll ask of any new table, and the trip data you'll practice on.
Free course. Your Maestro, an AI, .
Someone sends you a table and a message: "Can you take a look?" It's one of the most common moments in data work, and almost nobody teaches you what to do in those first few minutes. That first look has a name: exploratory data analysis (EDA). This course gives you a method for it: seven questions you ask of any new table, in order, before you calculate an average or ask anyone for a report, whether that someone is a person or an AI.
Before you see the method, place a bet: what would you do first?
Check yourself: Someone sends you a new sales table and asks for a chart of price versus units sold by the end of the day. What do you do first?
What a row is
Say what a row represents and check it with the column that identifies it. If a row is a trip, each trip appears exactly once.
Shape and types
How many rows and columns there are, what type of data each column holds and what period the dates cover.
What's missing, and why
Count the missing values in each column and look for their cause in the other columns. A gap is almost never an accident.
Each numeric column
See how each numeric column is distributed, with a histogram and its summary.
Categories
Count the values in each text column and look for stray labels or groups you didn't expect.
Extremes: error or real case
For each extreme value, look for evidence in its other columns before you decide whether it's a recording error or something that really happened.
Two columns at a time
Plot two columns together and describe the relationship, including the rows that don't follow it, without claiming a cause.
As you go, you'll keep a log: what you observe, the evidence behind it and what you need to ask whoever knows the data. At the end of the course you pull almost all of it into a small project that repeats the first six steps of the method.
If you already have a table waiting for you, tell your Maestro what it is and it will tell you where to start.
What you'll practice on
A table of 311 trips from Rueda, a bike-share system, from March 2 to March 15, 2026. Rueda and its trips are made up (synthetic data, CC0), but the table has the problems of a real table from work: repeated rows, gaps with a story behind them, labels that don't match and values that can't be right.
How it works
Each exercise comes with short Glossary term and Glossary term code already written. You run it, change it and read what comes back; you don't need to know Python by heart. Before many answers, we'll ask you to guess first. It's the fastest way to notice what your intuition took for granted.
If a line of code doesn't make sense, ask your Maestro right from the page.
This course teaches you to find and write down problems; fixing them is a different job, with its own decisions. Where it helps, we point you to the free Maestro courses that handle that.