A quartile divides ordered values into four equal parts. It is useful when a skewed distribution makes a single average a poor guide for choosing what counts as substantial activity. section 9.2.1
For total cycling traffic by location, quartiles turn a vague cutoff question into an explicit comparison: identify values representing the top 25% or top 50%, then choose a threshold appropriate to the decision. section 9.2.1
Filtering cycling locations
Half of the locations recorded 216 or more bicycles in total. With at least 10 years of coverage, that is roughly 10–20 bicycles per measurement date; whether to investigate that level remains a business judgment. section 9.2.1
cycling_totals.describe()
Use descriptive statistics on total counts to inspect quartiles before committing to a filtering cutoff. section 9.2.1
