Data Ingestion & Infrastructure

Dear Task mining experts, what are the best practices for Task Mining projects. Like do's and don't, how to start, how long to run data capturing, ......? ThanksBRTimo Peters

My SAP team added new fields to a table that I extract for process mining. My delta loads have been failing. Do I need to initiate a full drop and reload? Will that then pick up these new fields or is there something else that I need to do?

We are trying to generate a table which contains n rows starting from 1 to n where n is dynamic.  we tried the below code, but it creates only 9 rows from 1 to 9. //code insert into supplierNumbers (num)WITH RECURSIVE NumberSeries AS (  SELECT 1 AS value  UNION ALL  SELECT value + 1  FROM NumberSeries  WHERE value < 20 )SELECT value FROM NumberSeries; // is there any way to overcome this?

I have created a new space in studio to analyse master data changes in SAP, The data integration part is successful and the data model along with activity and case table also seems to load successfully. But when i use the same data model in studio, it shows that the tables have no rows.  Can anyone please help me. I have been trying to solve this since last week and still cannot figure out. :( What am i missing?  

I have created a new space in studio to analyse master data changes in SAP, The data integration part is successful and the data model along with activity and case table also seems to load successfully.  But when i use the same data model in studio, it shows that the tables have no rows.  Can anyone please help me. I have been trying to solve this since last week and still cannot figure out. :( What am i missing?

It is mentioned in the scenraio 5 of 'IN' Predicate function documentation page - IN (celonis.com)  Thanks in Advance!

Hi, I'm having issues with displaying the screenshots in the Task Mining Screenshot Viewer component in Studio Views.  The error is the following: 

Hello, can you help me with this? I want to see my screenshots in a View Sheet.

I am trying to create a variable and set a dynamic value to it using the below syntax. DECLARE @my_variable INT. But it shows syntax error near "@". I tried checking the syntax in the Vertica SQL and i could not find a difference. It would be helpful if someone would help on this or can share your prior experience on the same.

Dear All,Hope all is well. I would like to compare two event logs (one is Actual and Another Planned) for same activity.lets sayID; Activity; Planned Date111, "Activity1"; "15/10/2022"111, "Activity2; 17/10/2022" ID; Activity; Actual Date111, "Activity1"; "16/10/2022"111, "Activity2; 20/10/2022" Objective: I would like to know which activities are impacting my overall performance. Thanks.  

Do we refer to the end to end relationship between the Table B (1) and Table A2 (Case Table - N), by ignoring the intermediate tables (Table A)?

Hallo Team, I have been using the P2P dataset provided by the academy.  Had used the following statement to calculate the PU_median according to each case - PU_Median("LFA1", "_CEL_P2P_ACTIVITIES_EN"."EVENTTIME").  I have diffculties understanding the outcome (in pic attached) - Why do I get the PU_Median time calculated in seconds format?How can I convert the O/P format in seconds to DD-MM-YYYY ? - Had tried using {t } function, but it was an failureEven though there are different time stamps for each case wrt to activity - why do I get the same median time (1245079800000) for all the cases?  Can any I detail it . Thanks in advance.

Hello, I open this question that was already asked by another user 5 months ago and not answered from Celonis. Moreover, I'm not allowed to comment or ask for update on that post if the post is older than 30 days. I would like to query data in an Action Flow from the Knowledge Model and not from the Data Model. Currently I have an Action Flow connected to a Data Model and I have modules like Tools (set multiple variables) replicating the same KPIs I have in the Knowledge Model. If I need to update a KPI in the Knowledge Model I have to remember to update it as well in the action flow, this causes high maintenance and possibility of error. This feature is critical since one of the main goals of using Studio is the centralization of KPI definitions into a single source of truth, the knowledge model. If you have any update on this topic I'd appreciate if you can share it with the community. Thanks. BRAlejandro 

Dear Experts , I have P2P activity logs and i have users column along with it infront of activity who executed it. Is it possible to create a network diagram using network explorer with each node being users along with value of transactions they are executing ? Can you advise ?

I want to define the dimension so it includes activities in 2 conditions, that only has occurred before specific activities (ideally defined by LIKE "string%") OR upto specific activites. However, what I found so far for cropping methods (e.g. crop to null) requires specific name of the activity and doesn't allow LIKE predicate. Is there a way to crop activities to anything before specific condition / first_occ of specific condtion?

need to compare each of the the Part Name under Ticket.Id 123456 if any of them are the same in Ticket.Id 345691 and in the Ticket table (in the same row as the Ticket.Id) the list of Part names also found in the succeeding Ticket.Id Ticket TableTicket.Id Event time123456 8-Jul345691 9-Jul Parts tableTicket.Id Part Name123456 PartName1123456 PartName2123456 PartName3345691 PartName1345691 PartName3

For me, the output is not clear -   Given 'A' has 4 unique combinations (Munich - DE, NH, UK, TN) - as expectedGiven that 'B' has 0 unique combinations. As per my understanding, the O/P I expect is 3 (Berlin - DE, UK, TN)Given that 'C' has 1 unique combination (Berlin- NH) - as expectedGiven that 'D' has 0 unique combinations - as expected (all the unique combinations wrt Berlin - DE, NH, UK, TN are covered in the B and C Cases).Given that 'E' has 3 unique combinations (Bristol - DE, NH, UK) - as expectedGiven that 'F' has 2 unique combinations. As per my understanding, the O/P I expect is 1 (Bristol - TN) - as the combinations (Bristol - DE, NH, UK) are covered in the case of 'E' above.  Can you please correct me if my understanding of the concept is wrong? Also please provide your inputs. 

REMAP_TIMESTAMPS (celonis.com)

How would I expect this data to be visualised. Would it be as four parallel tasks or in sequence in record order. Does process mining support a parallel split?Case,Task,Time1,x,01/01/2023 00:00:001,y,01/01/2023 00:00:001,z,01/01/2023 00:00:001,a,01/01/2023 00:00:00

Dear all,I would like to have the same number of decimals for different abbreviated key figures.With ".3s" I can only define the total number of digits, but I would like to define the number of decimal places for an abbreviated number.With ".3s" -> 8.07k .... 17.0k but I want: 8.07k ... 17.01k  Thanks in advance! Best,Jonas

Variant and Process Explorer not filtered when using CALC_CROP in load script  How to ensure that Variant and Process Explorer are filtered when using CALC_CROP or CALC_CROP_TO_NULL in Load Script (Analysis/Tab/Component). When the crop is done through Selection Views, it works correctly, but when the crop is done using PQL in Load Script, filter does not work for Variant and Process Explorer. Examples below: FILTER ISNULL(CALC_CROP_TO_NULL(FIRST_OCCURRENCE['XXXXX'] TO CASE_END, ACTIVITY_COLUMN())) = 0 Or FILTER CALC_CROP ( FIRST_OCCURRENCE['XXXXX'] TO CASE_END, ACTIVITY_COLUMN()) = 1  Thanks 

For joining tables by timestamp I need to round by the minute. I found date_trunc listed as supported function but can't get it to work. update TABLEset MATCH_TIME_START_MINUTES = date_trunc('minute', TIME_START); gets me this: Execution error: ERROR: Function date_trunc(unknown, varchar) does not exist, or permission is denied for date_trunc(unknown, varchar) TABLE.TIME_START is converted with to_timestamp() before and then looks somewhat like '2022-02-20 01:04:32'. How can I "round_minute" in data jobs?

I have 3 tables:Fact Table A with n entries per case and a time dimension (round_hour)Case Table which containts the case keysAnother Fact Table B with m entries per case and a time dimension (round_hour)The data model is n : 1 : m. Unfortunately a generel snowflake model is not possible. Now I need to sum up the duration from Fact Table B aggregated by the time dimension from eventtime2 (round_hour) and display the sum in a chart. The time dimension for the chart is represented by the time dimension eventtime 1 from Fact Table A (round_hour). I tried bind, domain_table in combination with PU-Functions but nothing worked.  Is there a any way map those time dimensions from different fact tables with a n : 1 : m relation? 

Hi, I am receiving "Column name is reserved" for a particular column in table that is causing the complete extraction of that table to fail Is there any way to change the name of column while extracting ?

I want to display only the sum of all (claim cap) for Parcel status that were said to be lost. I will like to use a 'Number KPI'. Please can someone explain to me how I can go about it?