I think the root cause analysis in GUI (Conformance) only performs the first level of root cause, but is it possible to perform a multi-dimensional root cause analysis (second and third level) using Pycelonis? If so, please let me know how to do it.

Dear Sasaki-san,

Regarding second issue you said, could you check if celonis.datamodels returns list including your data model ? If empty list is returned, it may be your API key does not have appropriate permissions.

Regarding first question, I guess you will filter data by first root cause attribute, then try root cause analysis to find second root cause etc.

Best regards,

Kazuhiko

Hello Takata-san

Thanks for the reply.

I tried "Celonis.datamodels" as advised and realized that the id was wrong.

So it was not datamodel id = datamodel key, but something else.

Regarding multidimensionality, if I want to do what you advised (filter the data with the first root cause attribute and do a root cause analysis with the other attributes as multidimensional) in Workbench, do I need to write the code from scratch myself?

Or is it already available on Git hub or Pycelonis tutorials?

Dear Sasaki-san,

You can use ID or name in find method, so let's try by name at first.

Regarding multi-dimensional root cause analysis, I imagine I can use standard Analysis (Conformance) and use 'Load script' to filter data in Analysis.

If you find EKPO.BUKRS = 1000 is the first root cause, then you will set FILTER EKPO.BUKRS = 1000 in Load script and try second root cause.

Best regards,

Kazuhiko