Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
I have a set of tables that I am looking to create, but each table will have different columns based on the value of field 'groupname'. We have items with different attributes (color, profile, thickness, width, length, style, etc), but these attributes only apply to some of the groupnames. For instance, groupname A might only have values in color, thickness, and width, but groupname B might only have values in color, profile, style, and groupname C might only have values in thickness, width, and length.
I want to be able to do different tabs/tables for each groupname, so theres not a bunch of blank columns if that groupname doesn't have values in an attribute, but am struggling with how to control the data in this way. Is it a calculated dimension? calculated condition?
So groupname A table should be something like
item color thickness width
Groupname B table should be something like
item color profile style
Groupname C table should be something like
item thickness width length
I know I can build out the dimensions (item, color, thickness, width, etc) any way I want, its the controlling what items show up based upon groupname that I am struggling with. Also want to add that there are no calculations being done here, its basically just a data table returning values.
I think the creation of such multiple and quite similar tables makes more difficulties as it solved. I believe I would keep them within a table-structure like:
GroupName Item Attribute Value
a x color red
If this isn't really practicably you may consider to use a generic load approach like described here:
The Generic Load - Qlik Community - 1473470
- Marcus
Thanks for the suggestion, I'm using generic loads on some of the data already due to how our database is structured with this information. I just figured there should be some way to say I want to build a table with only data where Groupname is A. Then abother table with only data where Groupname is B. And so forth and so on. Seems to be an easy enough ask within a data set as to not have to create a bunch of different tables on LOAD.