Characteristic |
Beta |
95% CI 1 |
|---|---|---|
| hour_of_day | -0.17 | -0.40, 0.07 |
| day_of_week | ||
| day_of_week.L | -1.4 | -12, 9.0 |
| day_of_week.Q | -1.6 | -11, 8.5 |
| day_of_week.C | -0.17 | -9.5, 9.0 |
| day_of_weekE4 | 5.4 | -3.3, 14 |
| day_of_weekE5 | 7.8 | -0.30, 16 |
| day_of_weekE6 | -0.55 | -8.3, 7.5 |
| hour_of_day * day_of_week | ||
| hour_of_day * day_of_week.L | 0.15 | -0.55, 0.85 |
| hour_of_day * day_of_week.Q | 0.10 | -0.60, 0.80 |
| hour_of_day * day_of_week.C | 0.01 | -0.63, 0.64 |
| hour_of_day * day_of_weekE4 | -0.38 | -0.99, 0.21 |
| hour_of_day * day_of_weekE5 | -0.39 | -0.98, 0.17 |
| hour_of_day * day_of_weekE6 | 0.00 | -0.53, 0.53 |
| 1
CI = Credible Interval |
||
Model
Definiton
\[\begin{equation} \text{ride\_duration\_minutes}_{i} = \beta_{0} + \beta_{1} \text{hour\_of\_day}_{i} + \beta_{2} \text{day\_of\_week}_{i} + \beta_{3} (\text{hour\_of\_day}_{i} \times \text{day\_of\_week}_{i}) + \epsilon_{i} \end{equation}\]
The brm() model estimates how bike ride duration is influenced by the time of day and day of the week, including any interactions between these variables. This is done using Bayesian regression with a normal distribution assumption.
Parameters
The characteristics of the regression are:
- Hour of Day
- Day of Week - Linear Trend (L)
- Day of Week - Quadratic Trend (Q)
- Day of Week - Cubic Trend (C)
- Thursday (vs. Sunday)
- Friday (vs. Sunday)
- Saturday (vs. Sunday)
- Linear Trend (L) * Hour of Day
- Quadratic Trend (Q) * Hour of Day
- Cubic Trend (C) * Hour of Day
- Thursday (vs. Sunday) * Hour of Day
Posterior
