Forecasting Time Series Data with Prophet - Second Edition by Greg Rafferty

Author:Greg Rafferty
Language: eng
Format: epub
Publisher: Packt
Published: 2023-11-15T00:00:00+00:00

Figure 7.9 – Wolf population forecast with decreasing logistic growth

Prophet will predict precise decimal values and, of course, wolves exist in integer values, but this plot shows that somewhere between 2010 and 2014, the wolf population will die off. In a real scenario, it also matters greatly whether the last few remaining wolves are part of a breeding pair, and we have ignored that factor here.

Notice that because we have specified both a cap and a floor, Prophet has plotted both as horizontal dashed lines. When logistic growth is declining, even if there is no relevant cap, as is the case here, one must be included in your model. You may choose one that is arbitrarily high with no effect on your model, although be aware that it will be included in your plot and may make it appear as if Prophet’s forecast is very low.

You can, however, exclude it from your plots by including the plot_cap argument as done here: fig = model.plot(forecast, plot_cap=False), which modifies both the cap and the floor. Prophet does not currently support the exclusion of only one of these from your plot.

Prophet currently supports one more growth mode: no growth (or flat). However, the Prophet team is at work on some other modes at the time of writing, and those may become available soon, so keep an eye on the documentation. Let’s take a look at this final growth mode.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.