4  Interactive graph

In this section, we carry over our efforts to relativize the relationship between changes in temperature from 24 hours earlier and changes in taxi ridership from a week ago. Instead of presenting all of the data at once, however, we let the user target a specific temperature and see how ridership differs when the previous day was warmer or colder and by how much. The farther up the \(y\) axis, the more taxi trips taken “today” compared to “last week,” where each “today” corresponds to the temperature chosen by the slider, and we see the universe of last weeks when it was a different temperature.

Hence, for the default temperature, 20°, we see that if yesterday was around the same temperature, in general fewer trips were taken today. But if yesterday was much colder or much warmer, paradoxically to our general conclusions, more trips were taken today.

We have limited ourselves, as in the previous section, to post-Pandemic, rush hour traffic when the temperature was -5°–30°.