SQL & Data Analysis
26 skills across 7 areas, in the order you need to learn them. 4 of them end in work you hand in and we mark.
Covers the subject wellWe mark it against a checklist you can read first
Take the skill check — about 10 minutes26
4
~47
7
What you'll be able to do
- SELECT and projectionWrite a SELECT that returns exactly the columns asked for, with clear aliases
- Filtering with WHEREFilter rows with combined AND/OR conditions and get the precedence right
- Sorting and limitingReturn a deterministic top-N result set with an explicit, total ordering
- NULL semanticsPredict and control how NULLs affect a filter, a join and an aggregate
- Data types and castingCast between types deliberately and avoid silent integer-division and timezone errors
- Aggregate functionsAggregate a column correctly, knowing which functions skip NULLs
- GROUP BY and result grainState the grain of a result set and write the GROUP BY that produces it
- HAVING vs WHEREChoose correctly between filtering rows and filtering groups, and say why
- DISTINCT and duplicate controlDetect where duplicate rows came from and remove them at the right point in the query
- Inner joinsJoin two tables on the correct key and predict the shape of the result
- Outer joinsPreserve unmatched rows with an outer join without accidentally filtering them back out
- Join grain and fan-outJoin three tables at the correct grain and prove the totals were not inflated by fan-out
- Self joinsWrite a self join with unambiguous aliases to compare rows within one table
- Semi and anti joinsFind rows that do or do not have a match without changing the result grain
- Window function basicsAdd an aggregate to every row with a window function instead of collapsing the result
- Ranking functionsReturn the top N rows per group with the right tie-breaking behaviour
- Window framesCompute a running total or moving average with an explicit, correct frame clause
- LAG and LEADCompare each row to the previous or next row within a partition
- Date truncation and time seriesProduce a time series with no missing periods, including days with zero activity
- Cohort analysisBuild a cohort retention table from a raw event log
- Conditional aggregation and pivotingPivot a long result into columns with conditional aggregation
- Validating a resultCheck a query result for correctness before trusting it, and say what you checked
Where the proof comes from
The shorter route in
Questions people ask about this
Start on SQL & Data Analysis
Every skill above, ordered by what depends on what, with the projects above that as the evidence you actually did it. We work out where you already are first, so the path skips what you can do.
Start this path