Visualize random forest bike model predictions

In the previous exercise, you saw that the random forest bike model did better on the August data than the quasiposson model, in terms of RMSE.

In this exercise you will visualize the random forest model's August predictions as a function of time. The corresponding plot from the quasipoisson model that you built in a previous exercise is in the workspace for you to compare.

Recall that the quasipoisson model mostly identified the pattern of slow and busy hours in the day, but it somewhat underestimated peak demands. You would like to see how the random forest model compares.

Instructions

The data frame bikesAugust (with predictions) is in the workspace. The plot quasipoisson_plot of quasipoisson model predictions as a function of time is also in the workspace.

Print quasipoisson_plot to review the quasipoisson model's behavior.
Fill in the blanks to plot the predictions and actual counts by hour for the first 14 days of August.
- gather the cnt and pred columns into a column called value, with a key called valuetype.
- Plot value as a function of instant (day).

How does the random forest model compare?

Take Hint (-30xp)