🟣 Apply the Example to a Different Problem
Use the example's tools and techniques on your own problem and data.
Previously
This assumes you have successfully created a new project, the environment is set up, and the initial project runs.
Step 1. Confirm the Data
The different problem may use:
- the example's data, with a new question
- a different dataset
- data collected for this project
Whichever applies, record the source of the data and any license or terms of use. The source is credited in the README links.
If the problem uses different data, add it (or connect the data stream) following the example's project instructions for where data lives and in what format.
IMPORTANT: Do not commit secrets, private data, or oversized files. If the data cannot be committed, document how to obtain it and keep it out of Git.
Step 2. Identify What Must Change
Compare what the new problem needs to what the example does:
- the question being answered
- the data files, columns, or fields used
- derived values or settings that depend on the question or the data (value ranges, units, thresholds, window sizes)
- titles, labels, and units in outputs and visuals
Keep the example's techniques. Change only what the new problem requires.
Step 3. Adapt the Project
Make changes in small steps and run after each one. See Run Your Project. Fix one error at a time.
Step 4. Record Decisions
Keep notes on each analytical decision: what was chosen, what the alternatives were, and why. These notes support the Data, Techniques Applied, and Methods and Settings sections of docs/index.md.
Verification
- [ ] The problem is stated and differs from the example's.
- [ ] The data used and its source are recorded (e.g. in a data-card.md).
- [ ] No secrets, private data, or oversized files are committed.
- [ ] The example's techniques are applied to the new problem.
- [ ] Each change made for the new problem is recorded with its reason.