Statistics and Forecasting

Forecasting in the Tableau Desktop is a powerful tool to extrapolate the data and visualize its trends as well as confidence intervals. In this example, the main question is based on current data, how we could predict and forecast the COVID-19 cases in the upcoming weeks or months. In order to use forecasting in the Tableau Desktop, make sure you are on the Sheet 1 and DateTotal field is aggregated based on weeks: WEEK(DateTotal):

Now enable forecasting by going to Analysis --> Forecast --> Show Forecast:

The first visualization of moving averages of number of cases of COVID-19 on weekly basis with forecast for different counties is shown here.

Also, it is possible to look at the forecast summarization by going to Analysis --> Forecast --> Describe Forecast:

Finally, by going to Analysis --> Forecast --> Forecast Options, we could control how long we want to project or forecast our data, as well as choosing the model, etc.

Due to the visualizations of forecasting the number of cases of COVID-19 in four counties of Greenville, Pickens, Anderson, and Oconee, we see that the prediction for Greenville county has the highest variation or lowest confidence. On the other hand, other counties of Pickens, Anderson, and Oconee have higher confidence due to their lower variations.

The workbook to reproduce the visualizations in this section is shared here: https://bit.ly/3n7YY3R