THE RESULT
What the data shows
The largest computed metric is 0.8762 for 168.0; the smallest is -0.1437 for 12.0. Metric: correlation (Pearson correlation).
Seasonality and trend contribute to correlation. This descriptive use of all dates is separate from forecast feature selection.
| lag hours | paired hours | correlation |
|---|---|---|
| 1 | 17,303 | 0.8431 |
| 2 | 17,292 | 0.5923 |
| 3 | 17,285 | 0.4024 |
| 6 | 17,267 | 0.008699 |
| 12 | 17,237 | -0.1437 |
| 24 | 17,219 | 0.8192 |
| 48 | 17,180 | 0.6875 |
| 72 | 17,158 | 0.6416 |
| 168 | 17,075 | 0.8762 |
THE METHOD
From source to answer
Pairwise-complete autocorrelation on a complete hourly grid.
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
Use lag structure to propose forecast features and validate them chronologically.
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.