About rob_marshall

DavidSmith · ‎10-18-2019

ethancomstock@gmail.com Just to clarify on "Formula protection" What is it? Hard coded references to a list member SELECT, IF ITEM(xx)=, P1 Product.Cars Top Level Item is OK as that is structural (you cannot change the Top Level in a Deployed model) You can have CODE(ITEM(P1 Products))=, or NAME(ITEM(P1 Products)) because that is not referring directly to a list element. Still not good practice though!! Why? In standard mode, you will generate a rollback if you delete/amend the list member Anaplan doesn’t not allow end users to generate a rollback In Deployed mode, a "Production" list allows end users to delete/amend the list, so Anaplan “protects” the list against this Good Practice It is bad practice to have hard coded references I hope that helps clarify things David

shreyak.garg · ‎10-18-2019

Hi Preethi, The steps you have mentioned look identical to the issue we had. To my knowledge, it is still an unresolved issue. I will confirm with the support team if there has been further progress.

nathan_rudman · ‎10-17-2019

follow Rob's guide. Quick addition to it: - never include yourself (the person running the action) in the process - setting user to no access kicks them out, don't do it during a day's work

ssmakris · ‎10-17-2019

Thanks!

fredrickstraube · ‎10-16-2019

Hi Rob, This worked perfectly, thanks a lot! /Fredrick

AlejandroGomez · ‎10-15-2019

Hi Rob, I am glad you managed to put together a nice example with screenshots of using ANY as alternative. This example is really clear and meaningful. Thanks! Alex.

lejaon · ‎10-10-2019

I've actually had a couple of instances where I needed to employ this method. 1. If you're dealing with the text formatted line item, then creating a combination list (list 1 and list 2) like Alex mentioned is the best overall solution. 2. If you're dealing with a number formatted line item, or *list formatted line item, you can also use rank. *For the list formatted line item, the code has to be numeric or it has to be a numbered list or else, rank won't work.* This method is a bit klugey and requires a pre-built list for numbers but if you want to think outside of the box, I recommend trying it and see how this plays out. 3. Bonus point. I have been toying around with combining list 1 IFO and list 2 IFO and trying to find creative ways with combining these two using AND / OR and other summary methods. I haven't been able to find a workaround yet but if you do come across a solution, please do share with the community. Jay-

ArunManickam · ‎10-09-2019

It can happen in 2 scenarios, 1. If your source is having time and target don't have time. Please lookup appropriately on time dimension to fix it. 2. If one of the dimension is not having a top level in the source hierarchy. Thanks Arun

RobinSilk · ‎10-07-2019

I dug a little deeper and found that in both 1.5 and 1.6 I was using "RS - Level 2 Data Hub Start / 'SYS12 SKU Details'.Build SKU", but in 1.6 I need to be using "RS - Level 2 Data Hub Start / 'SYS12 SKU Details'.Build SKU". I think this might be the fix to the issue. The only problem is that the second Source Label is too long and I am receiving this error message:

Ryan8219 · ‎10-03-2019

Appreciate providing me with the contact @usman.zia

Ceri · ‎10-02-2019

Thanks for the prompt reply - this worked like a charm

Stig · ‎10-02-2019

Perfect, thanks @rob_marshall !

karank · ‎10-01-2019

Just found out what it was . It was my access 😛 all good Thanks

ArunManickam · ‎09-22-2019

Hello @HendrikDeCuyper , As David mentioned, we need to cross the selection (list formatted line item) and filter (boolean) with users, so that they are exclusive for each user. I am assuming you are applying user dimension for only these 2 line items and not the entire module. In this case, it should not normally consume more space. Other option is to consider is use "Quick filters" and personalized dashboards. Users can be trained on quick filters which are exclusive to each user and also these can be saved as "My page" personalized dashboards. Thanks Arun

catzert · ‎09-18-2019

Thanks. that's exactly the same to do it rather than trying to solve the mapping errors.

rob_marshall · ‎09-17-2019

@jkinnick , The best resource we have right now is the Planual which is fantastic. You can find it here: link Regarding subsidiary views, they are not bad in that they perform poorly, but often times when they are doing calculations, that logic is hidden away from other model builders. If model builders were to just create a new module, others will be able to find it, use it, and then not replicate the same logic in other modules. With that said, some subsidiary views are needed for filtering and removing dimensionality, but if they are being used for logic, please put the logic in another module. Hope this helps, Rob

luke_e · ‎09-13-2019

Amazing - really appreciate the detailed response @rob_marshall ! Based on your response and in the context of our model architecture, (1) and (2) are definitely the most onerous of the lot. It's also good to know in the instance of (3) that the evaluation is determined by the applicable cells/objects. The linked materials also appear to be directly related to our scenario so I'll give them a suss as well. And appreciate that for many of these scenarios, it'll depend; I imagine it'd rather subjective at times as well depending on the builder, but your input has definitely helped us in honing in closer to the crux of our performance woes. Thanks again. Cheers, L

rob_marshall · ‎09-12-2019

@kandasamy , Because some of the files may have been loaded as Private. To read more, please follow this link: Read this Thanks, Rob

rob_marshall · ‎09-09-2019

@somain.gupta , First, I would ask why you have all three models in the same workspace, usually the Data Hub will be in a workspace of its own for security reasons. Secondly, I would look at the actions to see which view the data is coming from and ensure it is the same as the first data pull in the first model. Lastly, how is the data being changed? Do the target modules have the same dimensionality? Thanks, Rob

greggil · ‎09-03-2019

Yep, I just found it. There were manually saved views it looks like that had hand picked the teams by segment for each of the dashboards. Not a good solution for us because new teams come/go monthly and this is a data management nightmare. Using the lists is much better since everything is controlled by user access to the hierarchy levels, and it basically accomplishes the same views but auto updates. Thanks for helping!

rob_marshall · ‎09-03-2019

@pvs.pooja , The ETL tool of choice will extract the data from the source system and, via a connector, load the data into Anaplan. Any transformation can be done within the 3rd party tool (Snaplogic, Informatica Cloud, Dell Boomi, or Mulesoft).

rob_marshall · ‎09-03-2019

@Alexandru.Rosca , To answer one of your question, resetting the numbered list to a "regular" list and then flipping it back to a numbered list will not reset the internal numbers. Let me ask you something, why are you wanting to reset the numbers as they really shouldn't be seen by the end user. Are you not using the code? I get you are using a create action by the end users (which does not offer the ability to create the code), but what I have used in the past is to update the code (where blank) in a different process. Sometimes I will create a "Save" or "Submit" process where the first thing it does is update the code of the list. Can you tell us a little more about the list and what it is used for? Rob

rob_marshall · ‎08-27-2019

@obriegr , Since the new workflow is not GA'd, I believe @VIGNESH.M was suggesting a solution that is available to everyone. Rob

Venky_Rogers · ‎08-27-2019

Hi Arun, thank you for suggestions, i will give it a try

alan · ‎08-26-2019

This was very helpful. Thank you to all contributors!

mikejorg · ‎08-26-2019

You could use filtering to remove unnecessary list items to avoid this error. However, @rob_marshall raises an an excellent point. Why would you want to export, change the data, then import? It would make more sense to configure the module and/or model to be friendly to user input.