Change Model End of Fiscal Year Setting while preserving all data in order to align week 53

[DavidKidd]

Due to a new controling entity, in our main budget model, we need to change our 'End of Fiscal Year' setting from 'last Sat in Jan' to 'Sat nearest end of Jan. 

Our current set-up gave us a week 53 in January 2016, but our new controlling entity requires us to align our financial calendar and budget a week 53 in January 2018. 

Updating our model time settings has the undesirable effect of removing all historical and forecast data, perhaps as the time periods are incompatible? Is it possible to update the model time settings without losing all model data?

[ChrisWeiss · Accepted solution]

Hi David,

Not sure if anybody has had success preventing that kind of change from removing all data (I have not). One option to consider would be to create a new model with your updated time settings, and then import the data from your current model into the new model including any necessary data transformations in either the source or target models.

Happy to provide more details if you'd like, hope that helps!

[DavidKidd]

Hi Chris,

Thanks for the reply. I think the option of creating a new model and doing a mass export/import is the path that we are going to head down, it's good to get some assurance that this is a sensible approach.

Regards

David

[DavidSmith]

Hi David

I’m assuming that the data in Wk 53 FY16 will need to be mapped (to Wk52 FY16 or Wk1 FY17). If so that causes a problem when trying to map model to model, because you can’t manually map time:time. 

So I think an export/import is the only solution:

For each data module:

1. Export the data to .csv
2. Find and Replace "Week 53 - FY16" to whatever the new week is
3. Create an import from the file back into the module

Once done for all modules, copy and archive the model (just in case!) and change the timescale

Then for each module run the import from the text file

If you need to apportion Wk53 – FY16 (e.g. split the value into Wk52  FY16 AND Wk1 FY17) that will require some additional logic

1. Create some additional modules (for each data source) to calculate the split of Week 53.  This will need to include the new aggregated values for the weeks either side of Week 53.  I would create line items with the labels for the target weeks
2. Export the data for the split
3. Create an import
4. Export the data from the original module as above
5. Create the import as above

Copy and Archive the model, change the timescale and run the imports from the original modules; Week 53 – FY16 doesn’t exist in the new timescale so the original data won’t come in

Then run in the data from the “split” export files and it will overwrite the original weeks that are part of the apportionment and replace them with the new value incorporating Week 53

I hope this helps.

David

[DavidKidd]

Hi David,

Thanks for the response.

I am going to map the data in FY16 Wk 53  back into FY16, in order to preserve the FY16 yearly totals.

I am going down the export/import  path. I have duplicated the model, changed the Fiscal year settings in the target model (which then deletes pretty much all module data), and am now importing data back into the target model from the previous model using the Anaplan functionality.

This throws out errors where there is a conflict on time e.g. where the FY16 wk53 does not exist in the target model. I am saving the .txt output and I will apportion this to FY16 in the target model at a later stage, which will be the tricky bit. Your suggestion is helpful and I reckon I'll be doing something along those lines.

Cheers,

David

[DavidSmith]

Hi David

Yes, Wk53 FY16 will be missing so those values will fail and will need to be incorporated into FY16

I'm glad the suggestion was helpful, even if time consuming!

David