identical type of attribute is separated in statistic view into two bulks, why?

Atilla

In my given dataset (training and test)  theres an attribute called department. This feature contains  2 categories  sewing and finishing - this two are obviously nominal types. In rapid miner on the 'Statistic' view the categories (sewing, finishing) of the attribute department are visualized as a bulk diagram where the category finishing is shown twice - separated into two bulks in the diagram. My question is: What is the reason that in the 'Statistic'  view the same category (finishing) is separated into two bulks?
Normaly I am expecting to see 2 bulks (sewing, finishing) but on the statistic view there are three bulks (sewing, finishing, finishing). Back on the 'Data' view I only see the feature department and its 2 categories but the statistic view displays 3 categories (see visualization) which I can do not understand why. Maybe I do not understand the visualization view or even the view is just incorrect but the dataset is right. So in the end maybe I just need to choose the right diagram in order to get an accurate view.

BalazsBarany

Hi!

Can you spot the problem?



I added some characters to the department name. You have spaces at the end of the department name. Use the Trim operator to clean these up.

Regards,
Balázs

BalazsBarany

Hi!

In this case I used Generate Attributes with a formula like: ">" + attributename + "<"

Regards,
Balázs

Atilla

This is awesome. I am just a speck away from understanding the way how you detected the spaces in the department name (>finishing <)?  Will try your suggestion about  the trim operator in order to get the variables(names) of the department trimmed. I appreciate your help 🙌