Calculate Demand Frequency and Save Co-efficient of variation in an output KF in IBP

[Former Member]

Hello Experts,

A couple of requirements from segmentation perspective:

1. How can I calculate the demand frequency for any combination. For example - If the storage level of the Sales KF is Month and a Product is ordered 10 times out of 12 in a particular year, then frequency = 1.2. Is it possible to achieve a similar result in IBP?

2. In XYZ segmentation we provide the threshold for segmentation (0.5, 0.7 etc). Based on this the system segments the products and update the attribute. The log of the segmentation jobs gives the count of the planning objects that it has segmented. Is it possible to get this co-efficient for each combination in an output KF?

Regards,

Suman.

[revanchatraban · Accepted solution]

Hi Suman,

1. Demand Frequency - There is key figure IO Average Demand Interval to indicate the demand interval. Additional there is attribute to indicate whether the demand frequency is frequent/intermittent or not enough data. This can be calculated using the Forecast error measure-Inventory Optimization. Please check the logic with manual calculations.

You can find the more details in the below thread.

https://answers.sap.com/questions/13166338/forecast-error-measure-when-to-use-mad-or-mape.html

2. It is not possible to store the calculated CV value for XYZ segmentation. As mentioned by Piyush in the above comment, Squared deviation in the results of forecast automation can be used for the calculation. you can try with one manual calculation and compare with the CV in forecast automation profile. The level for calculation for both segmentation and forecast automation should be same.

Regards,

Revan C.

[piyush_parekh]

Hi Suman,

1. Create a planning level at yearly level (say YEARPRODLOC). Create a calculated key figure(KF1) at same level as that of Sales KF (say MTHPRODLOC) with calculation - If Sales >0 then 1 else 0. Aggregate KF1 from month to YEARPRODLOC as shown below.

KF1 at YEARPRODLOC = SUM (KF1 at MTHPRODLOC)

Maintain calculation in the target key figure (frequency) as KF1@YEARPRODLOC / 12

2. It is not possible to get CV calculated by XYZ profile. You can explore Forecast Automation - Time Series Analysis functionality where one can see the CV calculated for each planning combination in the fiori app itself. If you want to capture CV in key figure, explore Manage Forecast Error Calculations app. Further details related to Forecast Error app are mentioned below:

https://help.sap.com/viewer/1a63df8166d448d3b286fecb01077b07/2011/en-US/1820bd572801bb38e10000000a44...

Hope this helps!

Regards,

Piyush

[Former Member]

Hi Piyush,

Thank you for the response.

On point 2 - The Coefficient of Variation (CV) - I understand that the CV is not visible while running the XYX segmentation job. I have a doubt on the option that you suggested about getting the CV stored in KF via Forecast Error app. The Error CV & the CV we use for XYZ to calculate demand volatility are different ? Then how will I get volatility via Error app? Please correct my understanding if I am wrong.

Regards,

Suman

[piyush_parekh]

Hi Suman,

You will find the formula for CV used in Error app. But I am not sure if both - Error App and XYZ profile use same mathematical expression in the backend. You can evaluate this for e.g. by creating a range of thresholds in XYZ profile like 0.1 for X1, 0.15 for X2, 0.2 for X3 and so on.

Regards,

Piyush