Improved Performance and Scalability on Visualizations
We continue to work on improvements to scaling of visualizations. So far, you can see that there is a new visualization action bar for zooming in/out and back to default:
On some visualizations you will also see better scaling as the size of a visualization decreases. For example, when reducing the size of the following column chart, the axis labels are removed. We will be continuing to make improvements in these areas.
Highlighting and Filtering in Display
Widgets in a display that use the same data set will now interact with each other when you make selections in an individual widget. If a data element (such as a bar representing a product line on a bar chart) is selected on one chart and the same element is projected (displayed) in another chart, then that same element will be highlighted on the other chart. If the same element is not projected in the other chart, that chart will be filtered by the selected data element. In the example below, when all visualizations are based on the same data set. Golf Equipment and Personal Accessories are selected on the bar chart in the top left and the treemap visualizations are both highlighted accordingly. The two summary charts do not display product line, therefore they are filtered.
Filtering in Display using Hierarchies
Using either the filter area in your dashboard, or a filter widget on your display, you will now be able to navigate a hierarchy tree for filtering, including making multiple selections from different levels of the hierarchy.
Better Decision Tree Navigation
Decision trees can be very busy and have many branches making them difficult to view. To allow users to better navigate a decision tree a minimap will be included in the top left corner. The area of the decision tree being displayed is highlighted on the map. Users can move the shaded area of the map around to view different areas of the decision tree. Users can also use the Top 5 and Bottom 5 buttons to add indicators for the most relevant branches of the tree.
Additional Insights on Predict Visualizations
Visual Insights will be included with more types of predictive visualizations. A new button in the visualization action bar is used for removing the insights. For example:
WASM Social Media influencers
Watson Analytics for Social Media will calculate an author influence score based on a combination of the number of posts by the author and the engagement with those posts. The Influential Authors report replaces the Top Authors report in WASM.
WASM Italian Language support
Watson Analytics will support documents authored in Italian.