THE RESULT
What the data shows
The largest computed metric is 1.529 for 3.0; the smallest is 0.4133 for 6.0. Metric: relative_change (Relative change).
Different weather, availability and missing-hour patterns remain confounders. Means use observed hours only.
| mnth | mean 2011 | mean 2012 | hours 2011 | hours 2012 | relative change |
|---|---|---|---|---|---|
| 1 | 55.51 | 130.6 | 688 | 741 | 1.352 |
| 2 | 74.29 | 149 | 649 | 692 | 1.006 |
| 3 | 87.73 | 221.9 | 730 | 743 | 1.529 |
| 4 | 131.9 | 242.7 | 719 | 718 | 0.839 |
| 5 | 182.6 | 263.3 | 744 | 744 | 0.4421 |
| 6 | 199.3 | 281.7 | 720 | 720 | 0.4133 |
| 7 | 190 | 273.7 | 744 | 744 | 0.4405 |
| 8 | 187 | 288.3 | 731 | 744 | 0.5418 |
| 9 | 177.7 | 303.6 | 717 | 720 | 0.7083 |
| 10 | 166.2 | 280.8 | 743 | 708 | 0.6895 |
| 11 | 142.1 | 212.6 | 719 | 718 | 0.4963 |
| 12 | 117.8 | 166.7 | 741 | 742 | 0.4148 |
THE METHOD
From source to answer
Month-matched means and observation counts, rather than unadjusted monthly totals.
The Python source exposes this study’s transformations. The complete project download includes shared preparation and evaluation routines.
THE NEXT DECISION
What follows from the finding
Investigate system expansion and service availability before attributing observed growth to behavior.
Where the conclusion stops
Historical system rentals measure realized use, not unmet demand or station inventory. Missing hours are unknown. Weather associations do not establish causal effects.
Related studies may reuse observations or holdouts. These are historical analyses; associations and backtests do not demonstrate commercial impact. Further model tuning needs new, untouched evaluation data.