Skip to Content

This document gives the step by step procedure to use R-Single Exponential Smoothing Algorithm in SAP Predictive Analysis.

R-Single Exponential Smoothing:


The R-Single Exponential Smoothing algorithm enables you to smooth the source data by reducing noise and performing prediction for the time series data by using R library functions. Time series data is a sequence of observations over a period of time.

Let us walk through a simple example to work on the R-Single Exponential Smoothing

Step 1: The below Excel file have the OrderID and Sales Amount

Note:

When you use a time series model for analysis, the model does not consider data from the selected data source. Instead, it forecasts by considering data that was used while generating a model.

Step 2: Open the Predictive Analysis Tool.

Step 3: Go to FILE-> NEW-> Select the Data Source (here in this case select Excel)

DataSource.jpg

Data.jpg

Switch to “Predict” tab to set up the forecasting.

Predict.jpgPredict1.jpg

You can create a model by saving the state of a trained algorithm.

  • In the application toolbar, choose the Designer button to switch to the designer perspective.
  • Expand Time Series sub type.
  • Drag the R-Single Exponential Smoothing algorithm onto the analysis editor.

  In the analysis editor, right-click the algorithm component and choose Properties

R-Single Exponential.jpg

In the properties view, perform the following steps:

  • Select Forecast as output mode, as you want to forecast the data.
  • Select the Sales Amount column as the dependent column. The algorithm forecasts the data based on the Airline Passenger column.
  • In the Period field, select Month (12).
  • Enter 2011 as the start year.
  • Enter 1 as a start period. As the period is Month (12), 1 implies first month of the year (January).
  • Enter 12 for the number of periods to predict.
  • Retain the default values for the advanced properties.

Predict2.jpg

  • The “Analyze the result” dialog is displayed. /wp-content/uploads/2013/08/2_271034.jpg
  • Select Yes to switch to visualization perspective.

To view the results click the Results and view them in Grid modeResult.jpg



Result chart.jpg

Then Visualize your results using “Visualize” Tab

Result chart2.jpg

Hope it helps to use R-Single Exponential Smoothing Algorithm in SAP Predictive Analysis.


Thanks,

Shivani.Yemula

To report this post you need to login first.

5 Comments

You must be Logged on to comment or reply to a post.

  1. Cihan Oruç

    I’d like to ask you a few questions. In this document years 2011 and 2012 are given as source datas, but they also appear in the predicted datas section. Moreover for 2013 all months have the same predicted values. I can not comprehend why 2011 and 2012 are predicted whereas we already have them in the source data file. Also, why are the all months of 2013 have same values?
    Regards.

    (0) 
    1. Onur Göktaş

      Hello Cihan,


      It comes from the nature of the algorithm; that predicted data and the source data appear both at all-wide data range except the first row, and the final prediction rows. For single exponential smoothing, you need only “level” parameter and the alghoritm can start calculations based on first value(observation) of source data as an initial value. For double and triple exponential smoothing; since trend and seasonality are considered (trend for double exponential, both for triple exponential) , predicted values won’t be constant like simple exponential smoothing and also you would notice that different number of observations are required to execute these algorithms(Blank predicted rows at start).

      About the same prediction values here ; it’s also coming from the nature of that alghoritm. The algorithm is based on observations and continues while the observations exist.

      We must recognize that as the lead time increases the forecasts will usually be progressively less accurate. We note that forecasts based upon the mean or the simple moving average also have the same property of being equal for all future periods. Of course, once the next observation is made, the common forecast value changes”*.

      * http://www.lums.lancs.ac.uk/files/mansci/18735.pdf)


      Regards,


      Onur

      (0) 
      1. Mario L

        Hello Shivani,

        Can you help me choosing the right algorithm in SAP Predictive Analytics.

        I think it is Time series algorithm, but i have 2 sets of data and i cant apply this data on Time series as it has only one field for prediction.


        I have a problem in choosing the right algorithm to predict 6months of data in sap predictive/Expert analytics.

        I have 2 data sets, with 1 data set having 12 Months of data for “Lead Time Days”.Another data set having 12 months of data for “Parts Usage Days”.

        we need to predict safety stock and Reorder Point for next 6 months with reference to current stock.

        At present we are calculating them in excel sheet. But we want it to be calculated in SAP Predictive Analytics.

        Regards,

        Sudheer

        (0) 
        1. Antoine CHABERT

          Hi Sudheer,

          Could you please refer to SCN rules of engagement here?

          The SCN Rules of Engagement

          Point 14

          Don’t resurrect threads. Don’t reopen a question that has already been marked with a Correct Answer.  If your problem looks the same and the posted solution does not work in your situation, then you have a different problem and you need to create a new Discussion referencing existing thread. You are allowed to “resurrect“ a thread only when the link in the answer is broken and you want the author to fix it.


          Kind regards


          Antoine

          (0) 
  2. Antoine CHABERT

    Hello Shivani, is this still working with the latest releases of SAP BusinessObjects Predictive Analytics? We are planning for content migration and we would like to make sure that the content is still applicable. Thanks & regards Antoine Priti Mulchandani

    (0) 

Leave a Reply