Skip to main content

I actually found a solution for a similar question here.

Proposed was the following formula:

 

FILTER "xx"."yy" =

CASE WHEN <%=option1%> = 'a' THEN 'value'

WHEN <%=option1%> = 'b' THEN 'value2'

END

;

 

However, as a whole list of purchasing groups needs to be excluded instead of '=' or '!=' we would need to work with NOT IN / NOT LIKE... Meaning something like:

 

FILTER "xx"."yy" NOT IN

CASE WHEN <%=option1%> = 'a' THEN ('Value1', Value2', ...)

END

;

 

But somehow I don´t get the syntax, right? Can anyone help?

 

Thanks a lot :)

 

 

Hey Nicolas,

if I get that right, we had a similar issue.

We did a little workaround and created in the Transformations a new table column which had the commodity groups prefiltered.

I think if the purchasing groups you wish to filter by button click won´t change this is a good solution.

 

Hope this helps. Also, its a nice column for future analysis and so on.


Hi Nick. I believe Daniel's solution will work.

 

But, here's another idea:

You could create a button dropdown that writes values to a variable which contains the filter. For example, if you create a button dropdown with option 'a,' then you will set the action to write something like "FILTER "xx"."yy" NOT IN ('V1','V2',....)" to a variable called <%=filter1%>. What you will then do is simply write <%=filter1%> into the load script. Just make sure you also have an option in the dropdown for 'no filter' and set the value of <%=filter1%> to "/*** Comment do nothing ***/"


Hi Both,

 

thanks for your proposals! Already tried the variable option which seems to work fine. That´s cool 🙂 Just wondering whether the transformation table approach is more scalable (we have quite some sheets).

 

However, for that one I am not sure whether I fully understood it. Could you explain a bit more Daniel?

 

Thanks!


Hey Nicolas,

 

let me try.

Initial situation:

  • we have several company codes (BUKRS)
  • in the past all had the same Commodity groups.
  • We simplified the commodity groups for one BUKRS
  • the old commodity groups are used by the others.

 

We wanted to select the commodity groups ass. with the one/our BUKRS.

The new Groups where alphanumeric instead of numeric.

 

So in the transformations we created a column with identifies the wanted groups. Also clustered in direct and indirect materials.

so the Syntax is a simple CASE WHEN

 

So with this culomn we can select, cluster and filter in all celonis programs (view, analysis, action flow, action engine). Which is a huge benefit over the variable option.

 

Hope my explanation was/is understandable.

If not, just ask.


Hi Daniel,

 

okay, I guess I got it now... Will try and come back to you in case of questions.

 

Thanks a lot, Nicolas


Reply