Sort & Turn On/Off Analytics
This guide will go through the steps to enable or disable particular analytics packages of Narratives for Tableau and sort those analytics in order of importance to adjust your narrative. These settings can be found under the "Analytics" tab.
To learn more, you can also watch Video Series: Analytics Settings.
Continuous Analysis Example
In our example, we are doing continuous analysis of two measures (sales and profit) by month. When comparing these two series, the narrative by default may include a comparison of volatility such as the last visible bullet in this example.
However, we may not care about that analysis, or we may want to move that analysis further up in the narrative. We can adjust this in the "Analytics" tab of our settings.
For continuous analysis, the analytic buckets that are executed by default are:
Segments: This looks for interesting spans within a series or compare interesting spans across series. Examples include peaks or troughs.
Trendline: This looks to discuss the overall trend of a series if it is statistically significant
Correlation: If there are multiple series (measures or dimension values), this looks for the degree of correlation between those series
Volatility: When noteworthy, calls out the level of volatility of an individual series and compares series by their volatility
If we wanted to disable the 'Volatility" analytic, we just click the orange toggle and we will see the status turn to "Off"
Or, if we wanted to keep it enabled, but move it up in our story, we can click and drag the volatility analytic to the top of the list.
In our example, we'll disable volatility. You will notice in the resulting narrative, there is no mention of how the volatility of the two series compares.
You can also toggle analytics for discrete analysis. The analytic buckets are Distribution, Clustering, and, for multiple-measure analysis, Correlation.
Percent of Whole Analysis
You can also toggle analytics for Percent of Whole analysis. The analytic buckets are Distribution and Clustering.