watsonx.ai

In Decision Optimization experiments you can create different visualizations for both input and output data. Sometimes data has several categories and you want to see the information aggregated or to see averages or other values.

You can create your own pivot tables in the Visualization view of a Decision Optimization experiment in watsonx.ai or Cloud Pak for Data.

For example, in the following Supply and Demand Planning model, the optimal decision values obtained for the delivery of products in the Explore Solution view is a table with 462 rows. Scrolling through this table doesn't make the comparison of different delivery decisions easy.

However, in the Visualization view, this pivot table shows the same information aggregated for each market region and region and product and month.

You can also see this information regrouped according to the plant where the product was produced, in this pivot table.

Or to drill down further, you can also see this information displayed for each month.

To create such pivot tables, in the Visualization view you can add and edit a pivot table which opens an editor where you can select your tables (input or output data tables), the columns, rows and options for the aggregation (sum, average, count, etc).

Example pivot table widget editor

In the JSON tab of the editor you can also view and edit the json code for this widget.

For more information, see the watsonx.ai or Cloud Pak for Data documentation:  Visualization widgets and Pivot table widgets.

#watsonx.ai
#Watsonstudio
#Decisionoptimization