Product Q&A

In my activities table, I have activities like set status and change status. I want to calculate what is second last status value for each case, the status values are given in the changed_to column next to the activity column

Hello Please send help! Im desperatly trying to count all cases with repair in time = 1 (true) If not repaired in time theres 0. SUM doesnt work, because there is more than one activity per case, so there are many rows matching the same case id in the data base. Im looking for the percentage of cases repaired in time (of total cases) carried out by one service point. My KPI: AVG( CASE WHEN "EventLog_UPLOAD_csv"."REPAIR_IN_TIME" = 1 THEN 1.0 ELSE 0.0 END) Unfurtonately, it counts rows, not cases. So the result isnt exact. Can someone help?

Hi All, my problem: I want to calculate the time between two process steps. These Steps are sometimes very often within the process sometimes never. Also they can be linked directly without any acitvity steps between them. Furthermore there is the possiblity that hundrets of process steps are between them. Additionally it is possible that both activities are repeated immediatly. E.G: Repeating Process Sequences like (A, B, C, D, B, D, D, C, E, F, , D, C, C, D) I would like to find the next D to every C and calculate the throughput time. Looking forward to your answers. Thank you! & Have a great weekend

Hi, We have an issue where images in analyses result in errors after transporting them between CPM4 environments. As we have a development and production environment of Celonis CPM4.6 we regularly move analyses and data models from the dev to the production environment. Analyses without images, or with the company logo, work fine. However, recently we encounter issues when transporting analyses that contain images. Is there any way to solve this without manually re-linking all the images?

Hello Celonis-Team, As an extension to my suggestion: Click here A full text search over all transformation scripts would be very helpful. This way you can quickly find out in which scripts a certain table, field etc. is used. Thank you Thomas

Unfortunately searching the Community and internet didnt get me the answer to this problem. Im trying to convert a value to date format with TO_TIMESTAMP, but the value seems to have the wrong type: LONG instead of STRING. Is there an easy way to convert LONG into STRING without changing the underlying table (i.e. involving the developer)? I see there is a function for STR_TO_INT, but not the other way around. My Celonis version is 4.5.0.

Hi All, Im try to get all my data model list in my celonis application, for the purpose migrating to prd to UAT etc is there any syntax in python to get the list of DM. Thanks,

I need to create a filter for Orders on an OLAP table where the delivery note was created in the last 48 hours, and I am struggling with it. Should I use EventTime or something else? (I am new to Celonis, so any advice helps) Thanks!

Hi, I have my automation KPI: AVG(CASE WHEN PU_COUNT(AR_Cases_csv, AR_Activities_csv.USER,AR_Activities_csv.USER IN (<%= AutoUsers %>) AND AR_Activities_csv.TASK NAME IN (<%= AutoTypes %>)) = PU_COUNT(AR_Cases_csv, AR_Activities_csv.TASK NAME, AR_Activities_csv.TASK NAME IN (<%= AutoTypes %>)) THEN 1.0 ELSE 0.0 END) Thats basically considering specific activities carried out by automation users. Now I would like to filter the process where these activities are not carried out by the automation users. Basically I need the negation of this formula, converted as a component filter. How can I do that given that component filter does not like aggregation function?

Good day, I need to set a filter at column level to get the release date of the PO.This would be on a TABLE OLAP and the filter does not have to be at component level.In case I do it in a single table, I applyFILTER _CEL_P2P_ACTIVITY_MASTER_DATA ACTIVITY_EN = 'Release Purchase Order '.I imagine that a PU_FUNCTION is used to set the ROUND_DAY("_CEL_P2P_ACTIVITIES". EVENTTIME) that affects only one activity.Thank you very much.

Hello Celonis-Team, In order to optimize the data jobs and to increase the speed, I try to select only needed table columns und filter table extraction to only needed data sets. Unfortunately, this is not easy because it is hard to figure out which columns are needed where. Therefore, the following proposal: Implement a Where-Use List. This should show which tables and columns are used in which filter and in which transformation, e.g. a indicator in the Extraction Table Configuration (per table, per column)By clicking you can see in which transformations the table or column is used. where_used986443 22.6 KB Also, it would be very helpful to see the execution time per task and step in the job log as list column. At the moment I have to open the tasks and steps one by one and calculate the times manually. execution_time1630511 21.7 KB Thank you Thomas

In OLAP table series color mapping, we need to set the threshold value as a value from a KPI or variable. Like for the one we can use the fixed value as the threshold value in the Color Mapping

I cant save overview in .pgn format. option does not appear: Export: BPMN export (XML) and export component (PGN)

Hi team, Our group has been recently migrated to IBC. Now i am trying to access the analysis and datamodels of IBC via python from my local system. But i am unable to login . I was using python in the old environment , and i was using API_ID and API_token, now which API_ID isnt availble for me. Is API_ID required to login? or is just API_token along with the URL is enough? Please suggest Code that i have tried: from pycelonis import get_celonis login = {celonis_url: my_url, api_token: my_api_token} c_source = get_celonis(**login) Thank you.

Hi All, I have the following requirement: Suppose a process is as Start->A->B->C->D->E->End. Now, I would like to view the Avg Throughtput Time for Activity E. There are cases where the process flows from A->E, B-> E, C->E and D->E. (different variations) So the logic should then calculate the Average of the Time Differences between Activity E and the activities which occur just before Activity E. (could be any one of A / B / C / D). We dont have the Endtime field. (it wouldve then been just a diff b/w Starttime - Endtime) I hope the requirement seems clear. Your help would be highly appreciated. Thank You.

Ive a table component with 2 dimension:Dimension 1: Combination of Company code + Vendor ("'BKPF".BUKRS||BKPF.LIFNR)Dimension 2: Count of documents (COUNT(DISTINCT BKPF.BELNR))However Im unable to filter the Document count with count > 1. The system returns the error An aggregate is not a function. Please check that there are no dimension and aggregations used together as function inputs.Can anyone suggest how to achieve the filter?

Dear all, I have a few different questions about the Process Data Model. I think there have been a few CELONIS adjustments in the past. Even though the relationship between activity and case table is correct (Process will be shown in the process explorer in the correct way) the display 1:N is mirrored (pic #1). In the past it was possible to declare/identify the case table with a . I cant find this option any more in the drop down option (pic #2). In general we need to insert two sources (different systems) in the same process analysis. They are running in parallel and are differentiated by additional information in the case (pic #3). The data is written into the case table. Difference than usual (normal create case table syntax), we use the SQL Vertica insert into syntax like we used in the activity table. We have to match two different systems with different data sources. In my opinion it will work fine, but I am not really sure that we can use it that way. Although I only create

Hi all, We have 2 (on premise, so not IBC) Celonis environments running: one where we develop our dashboards, and one production environment where finished and tested dashboards run. Transporting dashboards from DEV to PRD is only done by us, on request. However, were looking into automating the manual steps to reduce errors. Creating the transports from one or more dashboards and/or datamodels seems to be feasible through the Python API. However, Im wondering if the overwrite option is available in the Python API when uploading transports. The documentation does not mention this. Does anyone have any insights/ideas/suggestions?

Dear Celonis Community, I am having trouble creating a bar chart which tracks the amount of unsend orders. I need to filter all cases which do not have the Delivered activity and at the same time, have a preferred delivery date < current date. I managed to write two seperate statements but get errors trying to combine both. Maybe somone can help. SUM(MATCH_ACTIVITIES(EXCLUDING_ALL[Delivered])) and SUM(CASE WHEN TO_TIMESTAMP(CASES.Preferred_Delivery_Date, ) < HOUR_NOW(PST) THEN 1.0 ELSE 0.0 END) Kind regards. Konstantin

Im trying to convert a string field to date using TO_TIMESTAMP function in Celonis but receiving the error Undefined exception occured: ToTimestamp: Input does not fit to date format.The function works well for a constant value.e.g. TO_TIMESTAMP(20210101,YYYYMMDD)but does not work for a table field. e.g. TO_TIMESTAMP(CEL_AR_ACTIVITIES.CHANGED_TO,YYYYMMDD) returns error.Can someone help with the error?Thanks.

Hello! How can I put similar filter on a table? In KPI, I would write it in this way: CASE WHEN PROCESS EQUALS END THEN document deleted ELSE valid document END (means = Deleted) but how do you write it as a filter?

Dear all,I am encountering curious changes of results depending on the number of KPIs and Filters. I put an identifier in EKKO-VERKF for differentiating process patterns. Each pattern goes through maximum four, minimum two process flows. In other words, one CASE_ID contains from two to four activities.Here is the problem: the result changes after you add another KPIS, namely M04 disappears, from M01 to M03 gains/loses a few. I am wondering why one KPI affects the others, or rather if I write this code wrongly.Best regards,[KPIS]CASEWHEN SUBSTRING(EKKO.VERKF, 3, 3) = M01THEN CALC_THROUGHPUT(FIRST_OCCURRENCE[A] TO FIRST_OCCURRENCE[B],REMAP_TIMESTAMPS(ACTIVITIES.EVENTTIME, HOURS)) - 7*24WHEN SUBSTRING(EKKO.VERKF, 3, 3) = M02THEN CALC_THROUGHPUT(FIRST_OCCURRENCE[A] TO FIRST_OCCURRENCE[B],REMAP_TIMESTAMPS(ACTIVITIES.EVENTTIME, HOURS)) - 14*24WHEN SUBSTRING(EKKO.VERKF, 3, 3) = M03THEN CALC_THROUGHPUT(FIRST_OCCURRENCE[A] TO FIRST_OCCURRENCE[B],REMAP_TIMESTAMPS(ACTIVITIES.EVENTTIME, HOURS)) -

Hi, thank you for dropping by this topic. Im trying to use KPI Format as Start of MOVING_SUM. MOVING_SUM(COUNT_TABLE("_CEL_SD_ACTIVITIES_CASES"), KPI("Number of process variants") *-1,0) This KPI returns me an error: that start parameter should be an integer, which I firmly believe I am giving it integer. Using COUNT DISTINCT instead has not solved this situation btw. This should show the same result as the formula below. MOVING_SUM(COUNT_TABLE("_CEL_SD_ACTIVITIES_CASES"), -2,0) Best

Hi Team, I am working on uploading the file data to build eventlog in IBC. In the file, one of the file holds seconds and I would like to convert to time as hh:mm:ss, after that I need to add this time to existing timestamp to get the correct Eventtime in event log. Example : Field 1 Field 2 625 2019-11-20 13:41:25 I need to convert field 1 (625) to Time format hh:mm:ssthen add the time (in step1) to Field2 , this is not only adding, date should automatically change if the time exceeds 24 hrs I tried with the possibility of using ADD_Seconds function but it throws message as the function does not exist. Your suggestions are welcome!add_seconds1313228 10.6 KB

Dear all,I would like to create an OLAP table with the activity as dimension. And I would like the following KPI as a column:If all activities with the name of the activity in the dimension ran automatically for a sales order item, then count 1, otherwise count 0.For example:So if the activity Change Price occurs 3 times for an order item, and all price changes are delivered automatically, then count 1.But if at least 1 price change was manual, count 0. The following code does not work unfortunately. I suspect that is because the dimension is on activity level, and the calculation on item level.2020-12-22_11-09-13761117 28.7 KBSUM(CASEWHENPU_COUNT(VBAP, _CEL_O2C_ACTIVITIES.USER_TYPE, _CEL_O2C_ACTIVITIES.USER_TYPE NOT IN (<%=AutoTypes%>)) = 0ANDPU_COUNT(VBAP, _CEL_O2C_ACTIVITIES.USER_TYPE, _CEL_O2C_ACTIVITIES.USER_TYPE IN (<%=AutoTypes%>)) > 0THEN 1ELSE NULLEND)But is there any way to calculate this?Many thanks in advance!Best regards,Jonas