Oracle Analytics Cloud has a range of new capabilities including a new chart type, the Butterfly Chart. This can be thought of as a variant of the bar chart and is useful when wanting to directly compare two measures side by side.
As a simple example, I have created some sales v profit figures across years for some different products.
Using Oracle Analytics Cloud (or the newly released Analytics Server), I can quickly upload that tabular data and use a Butterfly chart to better see a visual comparison of the sales v profit data across the years.
So how does it compare with some other charts for showing the same kind of data? Let’s compare a Butterfly with a Bar chart and a Pivot table.
Starting with the bar chart on the left, I can see the figures side by side across the years and I have colour coded the sales and profit figures. Next with the Butterfly I have added an extra ‘dimension’ of colour for the product being sold. The Butterfly allows me to set a hierarchical colour which is indicated in the legend so that each Measure.Product in this case can be identified.
I duplicated the visualisation and switched it to a pivot table which you can see on the right – which now in OAC/OAS version 5.5 has the ability to add break and grand totals just like BI Enterprise Edition. With some colour coding this is quite readable.
In my opinion, having the native capability of the Butterfly does bring a really useful visualisation for adding clarity to the visual comparason of two measures and so it provides an opportunity to rethink any existing charts you have that may convey a clear picture to the consumer and re-engineer them.
Anything to look out for?
Negative values. Let us tweak the data to show how negative numbers are accommodated. I have updated all of the 2020 Profit values to be negative numbers. On a bar chart we would see something like this, with the 2020 Profit figures being negative:
If I flip that to a Butterfly chart we see no data representation for the profit in 2020. On the bottom right of the page we see a number of icons and we can see an Information Icon which indicates it is there to show Visualization Warnings.
If we toggle this setting on, then where there are data issues that can not be represented, then in the title of the visualization the same Information icon is displayed so that when we hover over it, we get a message showing the issue
Therefore, there are certain use-cases where Butterfly is perhaps not the optimal choice. Fortunately Oracle Analytics provides us with information to show that there is an issue with the data so that perhaps a different visualization can be selected. This is a really useful feature as of course data can change and what was perhaps the most appropriate way of visualizing certain data elements may also change as data volumes and data content change. As it is so straightforward to duplicate visualizations, experimenting with a range to find the most appropriate is an easy process.