Since SAP TechEd Madrid the announcement came that SAP BusinessObjects Predictive Analysis 1.0 is now generally available. Based on previous versions this current version is easier and simpler to install. It also now includes features from Visual Intelligence.
Please note that you cannot have Visual Intelligence installed on the same machine as Predictive Analysis
Also note the operating system requirements per the SAP Product Availability Matrix - it is 32/64 bit Windows 7 only and it only supports English as a language.
The link to the help explains a big change from the previous ramp-up version http://help.sap.com/pa
“SAP Predictive Analysis inherits data acquisition and data manipulation functionality from SAP Visual Intelligence.”
This is true from the first entry screen:
You can install R through the application – please see this blog/document for further details http://scn.sap.com/docs/DOC-33487
I thought I would follow the help on page 39 and use triple exponential smoothing against sample airline passenger data, to forecast the number of passengers flying in the future.
As shown above I've already acquired the data and "enriched" it using Visual Intelligence inside Predictive Analysis. I have "prepared" the data, now I will go to the Predict tab to set up the forecasting.
Then in the time series algorithms I select Triple Exponential Smoothing by double-clicking it
I configure the forecasting as shown above.
Then select the Data Writers tab and double click CSV output to output the results to a .CSV file
I configure the properties (showing the file name)
I "run" the analysis and receive the "Success" message:
To view the results click the Results and view them in Grid mode
Then Visualize your results using "Visi-like" functionality
Notice I had to change the PredictedValues into a measure. I did it a second time and it had "enriched" the values so I didn't have to change it to a measure manually. User error?
Above represents a visualization of predicted values.
On Share tab you can share the datasets as you would per Visual Intelligence.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
11 | |
11 | |
8 | |
7 | |
5 | |
4 | |
4 | |
4 | |
3 |