Data card — Old Faithful eruptions

How to load

faithful_df <- tibble::as_tibble(datasets::faithful)
# or: data("faithful")  # same table in base R

Where we use this

Caveats

  • Observational Yellowstone geyser records; not designed for prediction benchmarks.
  • Bivariate only — used to teach plotting and tidy data habits, not full modeling pipelines.

Data card

1. Source

Built-in R table faithful (also in datasets). Classic measurements of the Old Faithful geyser (Yellowstone): eruption duration and waiting time between eruptions. No separate download; no package beyond base R / datasets.

2. Outcome

No fixed outcome in our course — we treat both columns as exploratory. If you modeled: either waiting or eruptions (minutes) as a numeric target.

3. Predictors

Two numeric columns:

Column Meaning
eruptions Eruption duration (minutes)
waiting Waiting time to next eruption (minutes)

4. Sample size

272 rows. No missing values in the built-in table.

5. Leakage risks

Low for our use (plots only). If modeling: fitting and evaluating on the same 272 rows without a holdout would overstate performance; not a concern for the Day 1 visualization block.

6. Subgroups

Not used in class. In a research setting you might record season or measurement era; this table has no such fields.

Back to all datasets · Lab exercises