Data analysis: filter, group, aggregate
Almost every analysis you will ever write is these three operations in some order. Filter to the rows that belong in the question. Group by the thing you are comparing. Aggregate into the number you will report. Build a query here by choosing each step and watch the table change under it — including the ways it can quietly mislead, such as an average taken across groups of very different sizes. The syntax differs between SQL, spreadsheets and pandas; the operations do not. Learn the shape once and the tools become interchangeable, which is exactly the transfer the certification programs assume you already have.
About this tool
What it does
Runs the analyst loop on a bundled dataset: filter it, group it, aggregate it, and watch the table and chart change with every choice.
Who it is for
Anyone who has heard “just look at the data” and wants to know what that actually involves. No spreadsheet or coding experience needed.
Example
You enter:Filter to one region, group by month, average the sales column.
You get:A monthly average per month for that region, as a table and a bar chart — and the same dataset answering a different question the moment you change the grouping.
What it will not do
One small bundled dataset, chosen to be legible. You cannot upload your own data here, and the tool does no statistical testing — it is the shape of the work, not the work.
Your data
Runs entirely in your browser on the bundled dataset. Nothing is uploaded, sent or stored.
Next step
Data Analysis
Filter, group, aggregate
Raw rows rarely answer a question on their own. The analyst's move is to filter to what matters, group by a dimension, and aggregate a metric. Change the controls and the table and chart rebuild instantly.
| Group | Value |
|---|
Same data, different lens — group-by + aggregate is how one table answers dozens of questions.