About nathan_rudman

nathan_rudman

not exactly on subject but icons on a dashboard will block used from saving their views for "Personnal saved dashboards"

nathan_rudman

Anaplan has its own API, which can interact with any other. However you cannot write script inside Anaplan itself so you are going to need a server and a script to handle that. The easiest is a container with a node.js server where you can quickly write that script We also have a python library (look for it on community).

nathan_rudman

if my memory serves well the text of Anaplan is a very long text field of 3 millions characters

nathan_rudman

As Jared said, every time it is a little different. Personally, I'd recommend having your version comparison on a specific module where the versions are line items. The simplest is just to do a [SELECT:your version] for each line item. This give you complete flexibility on the logic (change the rate, replace fcst by actuals etc etc without touching the actual data. And it allows you to simply archive. Remove the formula and voila, your values are here to stay and you can re-use the actual version for planning. So yes, it's not fully dynamic and requires the admin to change formulas but in my opinion (having managed an FP&A model long term) it's the easiest for the admin.

nathan_rudman

similarly, you can reimport this data from the DEV model.

nathan_rudman

No unfortunately that icon is on the right. You can use quick filters as mentioned by T_Caban or you can published line items that you use as filter, so they look like a form on top of the module they filter

nathan_rudman

what's in EM41 ?

nathan_rudman

note that you can set an import in the dev model, importing from a copy (or even the prod model) into this list, sync. You can then remap the source of that import to the DEV model and run it to get the values back.

nathan_rudman

Hello Sameer, I think the best way to learn Anaplan is to build by yourself but I really recommend you do the basic training first.

nathan_rudman

Anaplan is a very flexible platform so the answer to that really depends on what you want to do with the numbers carry by the people. save the costs/revenue in the previous position ? transfer them in the new one ? have both ?

nathan_rudman

for me the main differenciator of Anaplan is its flexibility. In 5 years of doing Anaplan, I've been able to model 95% of what people had in mind , and I've been asked crazy things and very very diverse stuff.

nathan_rudman

is there another way than this to model in any technology today ?

nathan_rudman

I sent you a PM on the person you should contact at Anaplan as you seem to be an intgeration platform.

nathan_rudman

I'd add to what Misbah said: authentication method has no relation to import/export or anything you can do with the API once you are authenticated. Anaplan Connect is using the API thus can be configured to use certificate. Please follow this guide: https://s3.amazonaws.com/anaplanenablement/Community/Anapedia/Anaplan_Connect_v1_4_3.pdf

nathan_rudman

including the LOOKUP and the SUM ?

nathan_rudman

is one of the component of your formula pulling from a line item that as the the formula scope of Actual only ?

nathan_rudman

did you do anything ?

nathan_rudman

can you show us the URL ? I believe this is because you are building a "My Page" instead of a page of an app.

nathan_rudman

in the screenshot you showed us, the line items are not boolean formated nor time period

nathan_rudman

that's a clever one 🙂

nathan_rudman

RANKCUMULATE() also gives you some options

nathan_rudman

yes, if the admin modifies the main dashboard the view is reset.

nathan_rudman

yes it, if it is the same list (which it looks like it is) Edit the dashboard, click on the second module. Now on the left panel you should see the list of dimensions for this module. For your customer dimension, make sure "Synchronized Paging" is check and that should be it.

nathan_rudman

The difficulty is to find how many months you need to cover your AR days. For that, the only way in pure Anaplan is to calculate every possible way (another way would be to use the optimizer). Here's the result: and here's the formulas: You can see this is pretty standard. I'm using MOVINGSUM to calculate the sum of the past x months, compare that to the current AR, if it is higher, then I take the necessary days of the last months in the calc using LAG. DAYSINMONTH is very helpful here, giving us the numbers of days in the month. Now as I've said the trick is to know how many months we need. That's achieved with another module: You see I've created a list of items and I'm using their values to run the MOVINGSUM as many times as we have items. We then MOVINGSUM is higher than the AR, I return the value of the item, which is the number of month we need (-1 because inclusive of current). Then I set the summary method of that line item to MIN which is the result we want. That's why when the MOVINGSUM is below the AR I return 9999 so it is ignored by the MIN. Then we can bring this value into the "normal" module and use it for our calculation. The drawbacks here are: - it's not entirely dynamic as it relies on the items in the list. If you know that you average number of months is 13, create items from 5 to 20, depending on variability, and it should encompass all possibilities - this can be quite size heavy if you are doing it by multiple countries / products

nathan_rudman

Hi, Anaplan is an in-memory platform. When a model is first opened, it is loaded in memory, which means C is calculated. When you modify A, C recalculates, and also all other dependencies (but only those: if D = B, D isn't recalculated when A changes). This is the principle of Hyperblock, our patented technology. As this is happening in-memory, C isn't written to disk during this process. Obviously there is some write to disk events for data archiving and recovery, but I don't know if C is written to the disk then. @DavidS can shed some light

nathan_rudman

Hi, to reproduce this you should use the card named "Field" in the NUX. I can be one or several line items into one "field card"

nathan_rudman

yeah I missed that part 🙂 I have a question: what's the logic here ? I suppose you're going to have many more employees without a role than role un assigned, how is it going to be when all roles have been assigned once ? cycle back ? And the order at which these roles are assigned depend on what ? the order of the employees in the list ? Anyway, you can use the rank/rankcumulate formulas to count the employee without a job role, create a "placeholder" list with as many elements as employees without a role and apply that list to the logic I showed you. you can then add a [LOOKUP:rank] to find the right role. I can do that in a module once you answer my questions above

nathan_rudman

We have a university program in the US and in Singapore, and hopefully more soon. If you belong to these you can get access. If not, which university & location ? The local Anaplan team can work with you to start such a program.

nathan_rudman

you need to "extract" the un-allocated role, like this: The summary type of First UnAllocated must be "First non blank". As you can see, the total "All Roles" becomes equal to the first role that is not used as per boolean. That means any module without the role dimension pulling this line item will resolve on the top level "All Roles" Then in your employee list module, you can simply do: IF ISblank(Job Role) THEN 'Job Role:extract unallocated'.First UnAllocated else Job Role Obviously that works only for one default un-allocated role. You could add a dimension to the "First UnAllocated" line item if you want to assign several un-allocated roles depending on country or department for example.

nathan_rudman

I did not see your answer before posting. yours is valid and replies precisely to the "role not used problematic". However being able to count is useful as well, maybe it will help Mark. for the size, yes boolean is smaller. but here we are talking a module dimensioned by job roles only, maybe time, so extremely small. I used to do all this kind of counting by boolean, but very often you end up having to count as well, so then you need a number format, based on your boolean, which ends up being more space and calcs.