Product Q&A

Hi Celonis Community: I am following introductions of Use case: O2C - On Time Delivery Prediction from CelonisMLSuite. When I start training the trainer, the following error occurred. Could anyone tell me the possible reason for this error? image1369247 21 KB image1308227 17.5 KB Best Keith

Hello everyone, I am currently trying to set up a conformance analysis in Celonis. Thereby, I want to create a new process model for comparison. I have the new process model only available as a graphical version of the target process. However, as I am talking about an engineering process, return loops within this process are allowed and not a violation. Basically, I need the opportunity to allow some of those return loops as conforming processes. Whitelisting every violation case is in this case very troublesome. Therefore, I tried simply reconnecting the activities with arrows (see attached image) and gateways. Both do not seem to work. If anyone created process models like this or might have an idea for me, I am very grateful. Kind regards Matthias Ertl Example_New%20Process%20Model854260 6.35 KB

Hello all, how is it possible that the combination of calendar week and year is sorted correctly on my timeline? Starting with 1-2019, 2-2019, 3-201944-2019, 45-2019 image1338212 24.1 KB Many thanks! Best regards, Jonas

Dear fellow process miners, Im currently struggling implementing a normalization of a result in a table. It seems to be impossible to achieve that an OLAP table in Celonis. Heres my problem: Every case has a list of 0 to N events that affect the duration of the case. Its a simple parent-child relationship in the data model. If no event happened to a case, there will be no record in the child table. Now I created a simple OLAP table with the event category (event.category) as dimension. My two KPI columns so far are The number of events of that category as COUNT(DISTINCT event.key) The sum of delay in the process, caused by that event as SUM(event.delay). Now, I want to normalize the delay to 100 total cases. Say, I have 15 events, causing 15 minutes of delay. In total, I have 250 cases. What Im trying to achieve is the following simple formula: (SUM(delay)/Total number of cases)*100, which should be (15/250)*100 = 6. However, this se

Hi there, I work on an analysis to monitor, which sales order channel our customers use. Therefor I need a button for each order channel, where a viewer can select all customers ordering e.g. via EDI at on click. So the Viewer can analyse all customers with the possibility of ordering via EDI and can identify which other order channel those customers use. At the moment, the viewer has to select via a filter order channel EDI, then select all customers and remove the EDI filter, because otherwise the viewer can see only sales order items received via EDI. Can someone help me with the formula please? Best regards, Carina

Dear Celonis Community, I have a tricky statement to built and could not find a solution yet - maybe one of you have a nice idea! In our company we have so-called local customer hierarchy which is used on the one hand for grouping of customers and have settings on higher level than customer number. On the other hand used for reporting on different levels. This hierarchy is our a-level-hierarchy and it is usually used by the local affiliates. Technically the assignment can be found in entries from KNVH, where KNVH-HITYP = A. When reporting in headquarters a-level-hierarchy-numbers are grouped into f-level-hierarchy (also called GCN=GlobalCustomerNumber) and reporting is based on f-level. Technically the assignment can be found in entries from KNVH, where KNVH-HITYP = F. Basically as not always the lowest level of a-level-hierarchy is assigned to the lowest level of f-level-hierarchy, I h

We have some exciting news to share with all Celonis Snap users. Furthering our goal to democratize access to process mining, Celonis is taking the next big step, after the announcement of Celonis Snap, and makes all of its Online Training content for the Intelligent Business Cloud available free-of-charge. All Snap users now have access to our training tracks and can become Celonis experts today! This step-by-step guide will show you how to start your training. Log-in to your Snap and navigate to Training as it is shown in the picture below. 1600851 377 KB Then click Activate Online Training. 1600851 383 KB After that, you will be redirected to our training platform. Enter your details to sign-up, watch the introductory video and start with the training track that suits your role. 1600851 60.8 KB For instance, if you want to learn how to create advanced analyses choose the Analyst training track. 1600850 199 KB

Hi all, Currently we are doing an analysis on late Goods Receipt (i.e. Invoice receipt happens before Goods receipt). This is done based on the data model below: image.png1032482 32.3 KB However, now we got stuck on the issue how to calculate the time difference between the goods receipt and the invoice receipt. Using DATEDIFF(dd, MSEG.BUDAT_MKPF, RBKP.BUDAT) results in the error that these tables do not have a common parent in the schema. Any ideas on how to link these fields and calculate the delta in days? Thanks!

Dear Colleagues,I am trying to calculate average workload for a selected subprocess within the whole process.For that I wish to assess the average number of activities between specificed SOURCE and TARGET.It should show that from the moment invoice is received by AP accountant up till the moment invoice is posted, certain number of additional steps has to be executed. I wish to present it in various dimensions (for example in OLAP table by company code as dimension).I tried the formula presented in attached picture, but unfortunately it does not work.How can this scenario be applied?Thank you in advance for any suggestions.image.png957765 16.6 KB

Dear Celonis community, I wonder if you have an idea how to approach the following question: we want to define an approval limit for vendor invoices.to visualize the current distribution of invoices, we show the invoice numbers and aggregated values by grouped invoice value (as in the following bar chart prepared in MS Excel). Invoice_Value_Grouping.png1130744 35.3 KB Now, we would like to replicate this or a similar view in Celonis directly. Unfortunately, I cannot identify a standard Celonis functionality to group invoice values into n clusters as dimension. (The existing bucketing function is not flexible enough.) Question: did anyone already face a similar question and can share a way how to approach this? Any input would be much appreciated. Thanks, Jochen

Hi All,I am looking to build a case when statement in the load script area. Basically, I am building an analysis where certain local markets want to exclude fields . For example, for the Germany local market, we would like to exclude all FI vendors starting with 3 and in Spain they want exclude all real estate invoices, and so on and so on.As these filters will apply to the entire spreadsheet, I was hoping to write the code into the load script area, for ease of use, instead of each olap table but I am getting an error. Is this because you cannot write case when statement in the load script setting? And if not, can you please propose a workaround?Thank you,Anna

Dear Team, We are trying to create a dashboard with different KPIs. We have a scenario that we would like to filter in cuirrent status of the order. When we try to use Case ends with then there could be still some activites happening after that status and thus we dont see the results. Also, we would like to filter as an OR condition on multiple activities. FILTER MATCH_ACTIVITIES ( _CEL_O2C_ACTIVITIES.ACTIVITY_EN, NODE_ANY [ CRM Order: Released, CRM Order: Open, CRM Order: In process ] ) = 1; FILTER MATCH_ACTIVITIES ( _CEL_O2C_ACTIVITIES.ACTIVITY_EN, NODE [ CRM Order Item: Contains Errors ] ) = 1; FILTER CRMD_ORDERADM_H.PROCESS_TYPE in (ZTA, ZFD); In nut shell we want all the orders which are in Open, In process or released and contains error for certain process types. We are missing on how to apply filter on activities? Can someone help here? Thanks Naveen Gupta

Now, go to the analysis and start with the step On-Time Delivery Prediction 1/5 How to set up this particular Due Date and all?? Can anyone please publish any end to end Machine learning using python implementation in cellonis?

Dear all, I would like to measure throughput time between arbitrary two activities. I tried to use standard Process Overview sheet, but it does not work. Below screen is sample of demo environment. I believe I can see histogram of throughput time in the red rectangle area. Is there something tips to use it ? .png1594563 28.6 KB Best regards, Kazuhiko

Are you interested in trying Snap but dont want to upload your own data? With the App Store in Snap you will find a lot of demo data that you can install in your team and start analyzing. At the moment the following apps can be found in the App Store: Accounts PayableOrder to CashPurchase to PayService RequestsAccounts ReceivableUiPath Food Order ProcessSix Sigma To install an app follow these steps: In order to see the content of the App Store click on the button App Store top right corner in Snap: 1.jpg18971057 259 KB When you are in the App Store navigate to the apps. Select the app you want to install. In this example, UiPath: 2.jpg19031007 257 KB Click install App: 3.png18971063 162 KB Select New Workspace, enter the name of the workspace e.g. UiPath Demo, click next: 4 (1).jpg19131071 185 KB Choose the data model UiPath - Bot Analytics - Food Order Bot.

Hi All, As example reflecting only 1 specific SO, we currently have the following situation: All deliveries/billings from SO line items 10, 20 and 30 have a date in Q1 or Q2, whereas the date of the delivery/billing related to SO line item 40 is in Q3. Only Q1 and Q2 data is loaded in Celonis, so the delivery/billing related to item 40 is not yet loaded. Now wed like to calculate the total billed quantity for this specific Sales Order. As expected, PU_SUM(VBAP, _CEL_O2C_CASES.INVOICE_QTY) returns 40000, 40000 and 31600 for so lines 10, 20 and 30 respectively. However, if we want to aggregate over SO instead of SO line item and use PU_SUM(VBAK, _CEL_O2C_CASES.INVOICE_QTY), it returns 120000 whereas we only expect 111600. Any idea whats causing this sum of 120000 and how it can be solved to show the aggregation on SO level of only the loaded delivery/billing documents? Thanks

We have a data structure with case, activity, customer and contact tables.A contact is linked to a customer (N:1),A customer is linked to an activity (1:N),An activity is linked to a case (N:1).A case can be executed for multiple customers, hence this relationship structure of customers linked to activities.The challenge/question is to count the number of contacts for a customer of a specific case.Bonus challenge: only select those contacts that fall within the case runtime (which are attributes in the case table).The problem were facing is the famous no common parent between tables could be found error, and we believe the even more famous PULL UP functions could help us, combined with a DOMAIN_TABLE possibly. However, the relationship is such that we cannot just access a parent or grandparent table, but need to go to the parent (case) of a sibling table (activity) of our child table (contact), through customer which is our direct par

Hi All I would like to limit the scope of an anylsis by filtering some company codes in a load script. I have create a variable as text replacement in which I have put the list of company code: Variable Name= ShopCC TexT/Replacement= CH01,IT13,FR02,FR08 the load script is FILTER EKPO.BUKRS IN (<%=ShopCCode%>); Unfortunaltly I get an error saying that there iwas a syntax error in the load script. Any Idea please ? thank you

In a KPI card, I would like to display only the activities that occurred for that KPI for the last 90 days. I know the below is incorrect but logically it is as close as I can get without help. FILTER _CEL_P2P_ACTIVITIES.EVENTTIME BETWEEN TODAY() AND (TODAY() - 90) The KPI card works so I am looking for the code to tack onto to get the 90 day window. Please advise! Thanks.

Hello, Is it possible to create a stacked bar chart with Celonis ? Thank you for your help, Regards Robin

Im having some trouble understanding how the built-in Number of process variants KPI is calculated. I dont understand why these two statements are not equivalent. KPI("Number of process variants") versus COUNT( DISTINCT VARIANT( "_CEL_P2P_ACTIVITIES"."ACTIVITY_EN" ) ) image.png900323 21.4 KB

Hello all, In the PQL Editor an output of my formula is shown in the preview. If I then go to Done and want to look at the table, I get the following error message could not find referenced aggregate COUNT(DISTINCT ). Can anyone help me? I use the PQL formula: KMEANS() in its complex form. Best Regards, Patrick 01 Preview.jpg2156449 64 KB 02 result.jpg1383276 20.7 KB

Hi,I am trying to calculate payments terms in an Accounts Payable Analysis. Right now, my formula is:SUM((CALC_THROUGHPUT(First_OCCURRENCE[Vendor Creates Invoice] TO Last_OCCURRENCE[Pay Invoice], REMAP_TIMESTAMPS("_CEL_AP_ACTIVITIES".EVENTTIME, DAYS)))*(BSEG.WRBTR))/SUM((BSEG.WRBTR))Is it appropriate to do calculations combining BSEG and CEL_AP_ACTIVITIES Table? I would be curious to know if there is a better way to do this.

We are not able to edit any out of box analysis like Conformance checker, Social Analysis etc. We would like to use some of the components / formulas from these analysis into our single analysis for the users. How can we achieve the same?

Hi,I am trying to create a table that will show me which vendors operate in two different markets one of them being Spain. Some of vendors will be listed more then once in the market thus making the pull function a bit difficult to use I believe.Not sure where to go from here but Ive been trying different codes with no luck.Case when Cases.VendorName = Cases.VendorName AND WHEN Cases.LocalMarket = Spain AND Cases.LocalMarket = % THEN Cases.VendorName ELSE NULL ENDPlease let me know.Anna