The latest release of Oracle Analytics Cloud 18.3.3 and Data Visualization Desktop release 12.2.5 bring a wealth of enhancements, from data enrichment to new visualization types. The OAC release has some particularly interesting features, such as the replication capability, which is something we shall look at in a future blog post. In this short post, I wanted to draw your attention to some of the new visualization types that have been added in the September 2018 update. Let’s take a look at Correlation Matrix, Picto Charts and Heat Matrix Grid. Here’s a simple sheet of made-up data I created in order to show these. This sheet below shows some completely fictional data showing books in a Genre with an Author, Sales, Awards and Reviews.
Let’s load that up as a new Data Set
OK, let’s now create a new Project an a new ‘Correlation Matrix’ visualization.
What this does is to show the correlation between multiple matrices. This is expressed as a correlation coefficient from -1 (highly negative correlation) to +1 (highly positive correlation). We can see very clearly here that there is a fairly strong negative relationship between the Star Reviews and the Number of Sales. This is the effect of the skew we incorporated with the book “L” selling 200,000 copies and getting 1 in Star Reviews. Awards too of course also has a negative relationship.
If we remove book L from the dataset we see the effect of the deliberate skew.
Let us put Book L back and tweak the layout by bringing in Genre and using that as a Trellis column. This quickly and clearly shows that it is CELEB GOSSIP that is causing the skew.
Now if we flip that to the new Picto chart, this is a great visualization to show the distribution of data and here we are shoing sales by genre and author. The metric we are evaluating is ‘Sales’, so it is really clear how CELEB GOSSIP is driving our sales when comparted to the other Genres/Authors.
Let’s now flip to a Grid Heat Map. This very clearly shows that Book L on CELEB GOSSIP is the Sales champion !
Oracle OAC/DV already had a large selection of visualizations, and this version brings even more capability. Each visualization has a sweet spot based on the information we are analyzing and these new additions add even more analytic options.