Analyzing marked data with recommendations

You can mark data that you are particularly interested in, and then explore it from different perspectives. For example, you can view the details of the marked data in Details-on-demand, create details visualizations, or look for relationships to other variables in the data using the recommendations engine.

About this task

Note: The recommendations functionality described here is available only if you do not have access to AI insights. See AI insights for more information about analyzing marked data using AI.

Before you begin

The recommendations shown here are only available in the installed client.

About this task

There are several reasons you might want to explore the marked data by looking for relationships to other variables in the data. For example, you might want to examine whether there are differences between the marked data and the unmarked data, or find out if the marked data correlates to certain variables in the data. Visualizations that inform of such relationships can be suggested by the recommendation engine, and you can use these visualizations to get ideas on how to proceed the exploration to get further understanding of the data.

Follow the steps below to get visualizations showing relationships from the recommendations.

Procedure

In the visualization, mark the data you are interested in.
Click Recommendations based on marked data that becomes available in the visualization title bar. Alternatively, right-click the visualization and select Analyze marked data in the menu.

Note: The action initiates calculations in the recommendation engine. For big data tables, the calculations might take some time.
If the calculations discover there are relationships between the marked data and the unmarked data, these relationships are presented to you in the shape of various visualizations.

Note: Avg (average) is the default aggregation in the visualizations showing relationships.
Drag the visualizations, which contain relationships you want to examine closer, to the analysis.
A new column called 'Marked + a time stamp' consisting of the values 'Marked' and 'Unmarked' is added to the data table.
Explore the relationships in the added visualizations.

Tip: Some columns in the data table might be irrelevant to include in the calculations. Then you can exclude them before clicking Recommendations based on marked data . Open Data in analysis, right-click the column you want to exclude, and in the Include in recommendations drop-down list, select No.