About DavidSmith

Just adding, some of the checking could be easy, but other times, the calculation will have to run to know what the result is, but I agree that simple checks would be great ... View more

@jaddaffon  As @LipChean_Soh  and @JV271090  have outlined, it is nearly always possible to re-engineer the formulas to avoid SELECT. Along with the previous article posted, this is another good read: https://community.anaplan.com/t5/Best-Practices/Production-lists-and-ragged-hierarchies-logic/tac-p/37713   I would just add that until you do that, any changes to the lists will need to be promoted via a Synchronise.  Once the production model is in Deployed mode, you should never take it out!!   I hope that helps David ... View more

@elaine.novel  Ok, I think I understand what you are trying to do and it is a little involved and would require some maintenance if you create a new subset (and subsequent choice), but otherwise it is automatic if you change the members of the existing subsets   1. Subsets on the long list  2. Create a list "mirroring" the subsets   3. Create a module with a line item for each subset (dimensioned by the subset), and enter TRUE as the formula  4. Create a module dimensioned by the list in 1, and add a line item for each subset.  pint the line items to the respective formula from 3. 5. Create a module dimensioned by 2., with line items for each subset and manually check the respective element 6. Now we know which list items from 1 are in which subset and we know which list item from 2. corresponds to each subset , so lets bring it together in 1 module dimensioned by 1 and 2 with line items for each subset member and a check    7. I've assumed you have a module something like this, where the subset "choice" is made  8. You can now bring in the value for those list items that are in the chosen subset  9. This will then work if you change the elements in the subset   So the maintenance for adding a new subset would be: a. Add a new element to the list in 2. b. Add new line items in the modules from 3, 4, 5 and 6 c. Amend the check formula from 6.   Hopefully this resolves it for you David ... View more

One of the reasons why we advise against using list properties. Publish a module with line items instead David ... View more

@jbrass  Could you provide the business context as to "what" the purpose is for the ISFIRSTOCCURRENCE.  Why do you need it?  What is the significance of the time period? David     ... View more

@jensitseb  Definitely use a code rather than combination of properties if you can; it is much faster and more efficient. However, I would advise against using the month as part of the key.  If you have read my blog about sparsity: (The Truth about Sparsity (part 1), The Truth About Sparsity (part 2)) we have proved that in almost all cases, it is faster, and smaller to have a module dimensioned by time for the unique key (without month) and load the data values into that module Finally, we advocate not using list properties for the attributes, and calculating the attributes from the key in a module.   If you know the GL account is always 7 digits you can use the following to map directly again the GL account list GL account text = LEFT(NAME(Transactions),7) GL Account list = FINDITEM(GL Account, GL account text) This is more efficient for model opening as there are no text attributes stored If the GL account (or other attributes), vary, if you have a delimiter in the key, you can calculate the length, by using something like FIND("_",NAME(Transactions)) I hope this helps you and other readers stage data in Data hubs more efficiently David   ... View more

You can use Line item subsets rather than IF...THENs 1. Create a line item subset from the initial profile module and a module dimensioned by this LISS with the COLLECT() function 2. You can then set up a user choice module using the LISS formatted line item and pull in the COLLECT() line item against the user using a Lookup from the choice module 3.Then you will need to have the calc module dimensioned by user to pull in the profile from 2.   The calc module could end up being quite large if you have a large number of users   I hope this helps David ... View more

Unfortunately with column headers with time, if you don't map all of the non time columns it will show as an error as @aakash showed It would be better with two files, but as you said this is not an option, although, you can still add both actions into one process; the user would just have to run a single update, but choose two files as part of that process David ... View more

@ArunManickam  Firstly, you should always try and use a code for Numbered lists.   1. When importing 2. If using the create, you should still try and create a secondary import to populate the code   It makes it some much easier to manage later with a code   To your question, any import to a list needs to be uniquely identified, to know whether to add or update, hence why you have to specify Name, Code or combination or properties   David ... View more

@sarah.esguerra  Silly question Could you try bringing in the contract number as a number? and then convert to text to join with the company code Also, why do you need the ABS, that will only happen if the Company code is negative? And do you need the Company code and contract number as a number rather than a text field?   Finally, it would be better to provide the join in the source file if possible - that prevents all of the above! David ... View more

@Lakshman1997  No, unfortunately not,  there are a few different options in Time Settings though   This has been raised in the Idea exchange, so vote for it to bump it up the charts!! David ... View more

@usman.zia  Thanks! One of my favourite phrases is  "every day is a school day" David ... View more

@abhay.kanik  Yes, there is a simple way to solve this; I use this technique all of the time when I am trying to create codes and sub codes. You need to create line item in your module |(dimensioned by the numbered list). The formula is: RANK(1, ASCENDING, SEQUENTIAL, TRUE, Group By List) The last parameter is applicable is you are segmenting the list (by country for example), to ensure you have 2 number 1s for the different grouping.  If you only have 1 single list, this parameter is not required This will re-number any item when it is re-ordered or deleted You can then use this line item as the display name for the numbered list   I hope that helps David ... View more

As above, ensure that the formula is housed in your time settings module for optimal calculation efficiency David ... View more

Love this idea, specifically for True and False values.  Having to convert efficient Boolean values into numbers for conditional formatting is not great   ... View more

Good stuff - Glad it worked for you ... View more

Set up a module dimensioned by you "versions" list and create Boolean line items for each, so effectively it is Versions by Versions Tick each versions against each one, so effectively you have a "diagonal" set of tick boxes Then in you target formula you can refer to this module:   IF version mapping.IsversionFcst2? then else IF version mapping.IsversionPrevFcst then ..,. etc.   Does that make sense? David ... View more

@javed.bhat1  It is a manual option. When setting up an import definition in a DEV model you set it as production if you need the admin in the PROD (deployed) model to change the definition of the import. This is rare, as normally you shouldn't need to re-define the import action settings in PROD alone. More often the Import Source might change (e.g. coming from a different saved view), so setting the source as production is common.   Remember, if your model is in Deployed mode, don't take it out of Deployed mode.  IF you get an error if/when you try and amend the import, you will need to make the change in DEV.  Either make the change to the import and sync, OR, set the import to production and sync. DO NOT TAKE PRODUCTION OUT OF DEPLOYED MODE!!!!   I hope this helps David   ... View more