Capsules
First steps in Python and SQL, with AI
Twelve moves to read a data file, filter, calculate, make a chart, run a simple SQL query and get help from AI without copying blindly
- Non-developers: assistants, project managers
- Non-developers: assistants, project managers
- 12 capsules of about a minute
- 12 capsules of about a minute
- Video capsules in JupyterLite, FR and EN, with the exact code to type
- Video capsules in JupyterLite, FR and EN, with the exact code to type
- Beginner to intermediate
- Beginner to intermediate
What you will be able to do
- Open JupyterLite in your browser and run your first line of Python.
- Load a CSV file and look at its first rows.
- Keep only the rows you need, add a calculated column and a total.
- Total by group and turn it into a chart in one line.
- Save and download the result.
- Open a SQLite database, then filter, sort, count and add up in SQL.
- Ask AI for code, then check it against a number you know.
- Understand an error message with AI, and fix it.
The method
Everything happens in JupyterLite, on jupyter.org: free, no account, nothing to install. Python with pandas for tables and charts, SQLite for SQL. The code is shown as is, to type line by line; AI comes in at the end, to help, never to copy with your eyes closed. Data 100 % made up.
Module outline
Sequence by sequence, timed.
- 01py01 — Open JupyterLite and run your first line.
- 02py02 — Load a CSV file.
- 03py03 — Keep only the rows you need.
- 04py04 — Add a calculated column and a total.
- 05py05 — Totals by store.
- 06py06 — A chart in one line.
- 07py07 — Save and download the result.
- 08py08 — Open a SQLite database and read a table.
- 09py09 — Filter and sort in SQL.
- 10py10 — Count and add up in SQL.
- 11py11 — Ask AI for the code, then check it.
- 12py12 — Understand an error message with AI.
Capsules
Being filmed
Twelve moves to read a data file, filter, calculate, make a chart, run a simple SQL query and get help from AI without copying blindly
The capsules of this track are written and being filmed. The program above describes each of them.
The habits
- 1
Write the column name exactly as it is in the file.
- 2
The chart shows the same numbers as the table, nothing more.
- 3
Here, files stay in this browser. Download what you want to keep.
- 4
Never put real data into a public AI, and always keep a check number.
- 5
Understand the fix before you apply it.
The prompt library
- I have a pandas table df with the columns [your columns]. Write one line of Python that gives [the result you want].
- [the line of code] [the error message] What does this error mean?
- SELECT [columns] FROM [table] WHERE [amount] > [threshold] ORDER BY [amount] DESC;
- SELECT [column], COUNT(*) AS n, SUM([amount]) AS total FROM [table] GROUP BY [column] ORDER BY total DESC;
The resources
The resources
Sheets and materials to download or print.
First steps in Python and SQL, with AI
Non-developers: assistants, project managers
Newsletter
Monday's capsule
One skill a week, straight to your inbox. The easiest way to keep learning without thinking about it.
Your data stays in the EU. Never sold.