Hi Vclav,

In order to use the functions RCALL and RAGG a R server has to be set up and added to the Celonis configuration. This is currently still possible for Celonis 4 but is not supported in the IBC. We just updated the R-Documentation on our CPM 4.4 help:

https://help.celonis.de/display/PQLE/R+Server+Functions

. I also have attached the relevant documentation as a PDF.

As an alternative to R, you can also use our Python API.

The Python API works directly on the data model to retrieve necessary data, run calculations on it and push results back to the data model. Due to the vast amounts of available machine learning packages and a huge open source community for Python, you can apply any prebuilt code or scripts already available in your organization.

If you are currently using Celonis 4.3/4.4, then you can access the full documentation for the Celonis Python API by downloading it from

my.celonis.de

.

If you are using the IBC then you can find the documentation for the Celonis Python API at

https://[your_team].[your_realm].celonis.cloud/documentation/

I hope this helps,

Best Regards,

Calandra (Celonis Data Science Team)

PQL-R Server Functions-Documentation.pdf (3.3 KB)

PQL-RCALL-Documentation.pdf (9.4 KB)

PQL-RAGG-Documentation.pdf (7.4 KB)

Apologies for reviving an old thread, but I was wondering if there R will be

supported in the IBC at some point.

We are looking into running statistical tests (student T, Fischer, etc.) to check statistical significant correlations between an KPI/feature and an outcome (customer satisfaction). We would love to let our client play with this, and then have Celonis indicate whether that feature is relevant (for the current selection).

We are planning to use R for this (CPM 4.5), but are at the same time planning a move to the IBC. It would be a shame if this functionality (through R or another way) would not be available in the IBC.