
A Bootstrap Sample Can Repeat One Observation and Leave Another Out
Resampling with replacement can look strange the first time you see it: one original observation appears twice while another is absent. In a basic…

Resampling with replacement can look strange the first time you see it: one original observation appears twice while another is absent. In a basic…

A remarkably good model score deserves a question about the data split, not just congratulations. If the same examples appear in training and evaluation,…

A cumulative percentage adds shares as you move through an ordered table. Its final value is a useful check, but only after you know which observations…

A count rising from zero to five is a clear increase of five, but the usual percentage-change formula cannot express it as a finite percentage. The starting…

Inspect CSV quoting before counting visible lines as records, especially when fields contain addresses, notes or other multiline text.

Use an explicit, unambiguous ordering rule when extracting a limited database sample so first ten does not mean whichever rows happen to appear.

Read the quantity estimated by a confidence interval before treating its endpoints as the smallest and largest observations in a dataset.

Record a simulation’s seed alongside its settings and code version so a repeatable result is not mistaken for evidence that its assumptions are correct.