Definition
Could you now load a messy data set, clean it with pandas, store and query it with SQL, explore it with summaries and charts, and present what you found in a clear chart or dashboard?
This review brings together Working with Data: NumPy arrays, shapes, broadcasting and statistics; pandas Series and DataFrames; reading, exploring, selecting, filtering, grouping, merging and reshaping data; handling missing values, duplicates, wrong types, outliers, text and dates; databases and SQL with SQLite; and visualisation with Matplotlib, statistical charts, exploratory data analysis, data storytelling and dashboards, followed by a scored quiz.
In simple words
Read the recap table, try the activities, then take the quiz. Each answer comes with an explanation. If you miss one, go back to that topic.
Why it matters
Every AI project spends most of its time on data. These skills let you prepare data for models, check its quality and communicate what it shows.
How it works
| Topic | Key idea |
|---|---|
| NumPy | Fast arrays; shape, indexing, vectorised maths and broadcasting |
| pandas basics | Series and DataFrames; read_csv; head, info, describe |
| Transforming data | Boolean filters, sort_values, groupby with agg, merge, pivot and melt |
| Cleaning | Missing values, duplicates, types, outliers, text and dates |
| Databases | Tables, keys and SQL: SELECT, WHERE, GROUP BY, JOIN |
| Visualisation | Match chart to question; label axes and units; bars from zero |
| EDA | Shape, quality, summaries, distributions, relationships, groups; always plot |
| Communication | Action titles, highlight, annotation, honest context; dashboards with KPIs |
Try it yourself
Match the task to the tool
Which tool or method fits best?
- Average marks for each school
- Combine a students table with a schools table
- Count missing values per column
- See how a variable changes over time
- Compare distributions across groups
True or false?
Decide for each statement.
Data sets with identical statistics always look the same when plotted
Dropping all rows with any missing value is always best
Bar charts should start at zero
SQL GROUP BY is similar to pandas groupby
Common mistakes
- Modelling before exploring and cleaning the data.
- Charts without labels, units or honest scales.
- Losing track of cleaning steps so results cannot be repeated.
Check yourself
Summary
- NumPy and pandas load, clean, filter, group, merge and reshape data; SQL stores and queries it.
- Explore before modelling: summaries plus plots, groups checked separately.
- Communicate with honest, focused charts, clear stories and useful dashboards.
Finished reading?
Sources: pandas: chart visualization · Matplotlib: user guide and tutorials