THE RESULT
What the data shows
The largest computed metric is 0.3024 for lights; the smallest is -0.009289 for rv2. Metric: spearman_rho (Spearman correlation).
rv1 and rv2 are explicitly source-generated random controls, not physical measurements. The remaining energy and sensor observations are real.
| feature | spearman rho | absolute rho | paired rows |
|---|---|---|---|
| lights | 0.3024 | 0.3024 | 19,735 |
| t1 | 0.244 | 0.244 | 19,735 |
| t_out | 0.2155 | 0.2155 | 19,735 |
| rh_1 | 0.06052 | 0.06052 | 19,735 |
| rv2 | -0.009289 | 0.009289 | 19,735 |
| rv1 | -0.009289 | 0.009289 | 19,735 |
THE METHOD
From source to answer
Rank correlations for declared controls and selected real measurements; audit control duplication.
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 negative controls to question apparent signals before selecting features.
Where the conclusion stops
Measurements come from one home over a limited period. Energy is Wh per recorded 10-minute interval. This is not a representative household sample; tariffs and occupancy labels are unavailable.
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.