RapidMiner

Diagram: Being able to see instance data

Status: New
by MarcMiltenberge ‎10-01-2016 10:50 AM - edited ‎10-01-2016 10:50 AM

If you hover over an instance e.g. in a scatter plot, you can see the respective X and Y values. But it would be convienient to be able to navigate to the data editor with that particular instance selected on a double click (or via the context menu), so that one can see all attribute values of that instance.

 

Is there already a similar functionality?

Comments
RMStaff
RMStaff

Hi,

 

This functionality is already available but only for data sets with an ID attribute.  You can then double click on a data point, e.g. in a scatter plot, and a dialog pops up showing all the attribute values also for the dimensions which are not shown in the plot.

 

In case your data set does not have an ID attribute on its own, you can easily generate one using the operator "Generate ID".

 

Hope this helps,

Ingo

MarcMiltenberge
Contributor II

Dear Ingo,

thank you very much for this hint; I selected a particularcolumn (with unique values) as the ID column and it worked. On a double click, RapidMiner says "No visualization available for object with id K:\DataInput\...\a.txt" (the value of the id attribute).

However, this is enough information for me, since knowing the value of this particular attribute is sufficient.

 

By the way: Is there a way to specify within a CSV file that the role of a column is the label or ID column? I am setting these roles manually on import, but maybe there is a smarter way to save a CSV file so that the role of a column can be read in from that file.

 

Thank you very much for your help.

 

Best regards,
Marc

RMStaff
RMStaff

Hi,

 

Glad that the ID part worked.  On the role definition in CSV: no, unfortunately you can't specify roles in the file itself.  The CVS standard does not support something like this.  So you will need to select during import or - this is how I usually do this - I just import the data as is and then use the operator "Set Role" for this change as part of a data prep process I quickly run on that file post import and before I start with my analysis.  This way I have all the necessary preprocessing steps in one place.

 

Cheers,

Ingo