Skip to Content
Author's profile photo Tammy Powlas

Predicting the Future using a BEx Query as a Data Source

Using SAP Predictive Analytics (expert) I used a BEx query of actual expenses by project by month to forecast the future.  First I used real data but for the screen shots below I used sandbox (fake) data.  A use case for Predictive Analytics is time series analysis which can be used to forecast sales or expenses.  My example below is to forecast expenses.

After launching the Expert Analytics, I select Download from SAP Business Warehouse:


Next, I select a BEx Query.  It recognizes the prompts (variables) from the BEx query.


The above shows the following configuration for the triple exponential smoothing R algorithm:

1. Select the Dependent Column; the column of data we wish to forecast which for this is actual expesnes

2. Enter the Start Year and Period

3. The periodicity of the data as monthly

4. 12 Periods to predict

5. Leave the Advanced Properties default values.


The above shows the BEx Query (Project Cash) as the data source to the Triple Exponential Smoothing algorithm


After running the algorithm the above shows it was successful.


The above shows the predicted column on the far right.


This is one of the results views that comes with the Expert analytics package.  As I am a detailed person, I want to download the detailed results to a CSV file.


The above shows the properties of the output of the CSV writer.

I rerun the analysis and review the results in Excel:


The above shows the output in Excel.

More to come as I have time.

Assigned Tags

      1 Comment
      You must be Logged on to comment or reply to a post.
      Author's profile photo Abdel DADOUCHE (SAP)
      Abdel DADOUCHE (SAP)

      Hi Tammy,

      Soon, you will be able to use the "Auto Time Series" algorithms which allows you to include extra-predictable dimensions in your model.