Power BI needs to have a clustering analysis setting so that data queried via PowerQuery can be easily and automatically segmented based on one or a combination of columns containing KPIs. Or, this option can be embedded into a Pivot Table. When I say clustering analysis, I mean something like a K-means or Hierarchical clustering technique that uses statistics to analyze and group large volumes of data. If not these two methods, I have developed a technique using the IQR box & whisker plot rule that is simple to apply and operationalize. I am quite surprised that Data Analysis Toolpak does not even have a basic feature for clustering.
No CommentsBe the first to comment
Recent ideas
extend storage capacity of varchar(8000) to varchar(max) for warehouse
The maximum string length of 8000 is often considered too short to store longer descriptive data or embed a JSON value within a text field in a database. Dividing the data into smaller chunks and st...c_mettler22 hours agoNew MemberCompleted3.3KViews61likes13CommentsAdjust column width in SQL Analytics Endpoint user interface
I can't see the column names and cell values in some of the columns in my tables in SQL Analytics Endpoint because the column width is too narrow. Also, the tooltip also doesn't show the entire...frithjof_v22 hours agoCommunity ChampionCompleted4.2KViews75likes8Comments