First Steps from BW-IP and PAK towards SAP BPC 11.0, version SAP BW/4HANA planning with Embedded Model

[frank_klingl]

The last main set of new features of the planning engine was shipped with the SP4 of SAP Business Warehouse 7.5 (see SDN blog on ‘Planning Engine Improvements for BPC Embedded Model, PAK and BW-IP with BW 7.50 SP4' for details). I just recall the advanced DSO planning  as main innovation here.

Now we move our development activities in BPC Embedded planning from SAP BW toward BW/4HANA. In this step we also consolidate and simplify our offerings. Since BW-IP and PAK are clear subsets of BPC Embedded Model we will let them merge into the BPC4/HANA Embedded Model. The Embedded Model has two main flavours: the IT flavor where the classical planning objects (as also known from BW-IP) are modeled in the BW Modelling Tool beside other BW/4HANA artifacts, and the LOB flavor which uses manly the BPC Web Admin Client. Of course both can be used jointly and work seamlessly together. The very old BW-BPS based planning solutions are not supported in BW/4HANA. They have to be manually migrated to the BPC Embedded Model.

But our journey to make BPC Embedded faster, more flexible, and more user-friendly (?)  of course continuous also this year.

Main improvements are:

• PAK improvement for reference data selection in FOX


• Advanced DSO planning improvements and INT 8 key figure support


• Improvements in input ready queries and new lines on any level


• Characteristic Relationship maintenance in the BW Modelling Tool


• Improved Current Member support


• Input ready Dynamic Restrictions (created by end user)


• Dynamic grouping

 

  

Most of the features are also available in SAP BW 7.50 SP8 or the embedded BW with the SAPBW 7.52 software component in Business Suite applications. But some of the features are only available in SAP Business Planning and Consolidation 11.0, version for SAP BW/4HANA. In the future the shortly phrased BPC/4 will be the mainly focus for such improvements.

Let us have a glimpse on each of the topics in detail.

 

 

  PAK Improvements for Refernce Data Selection in FOX

 

You might remember that with BW 7.5 SP0 (see Improvement in 7.50) we introduced in FOX the ability to access data not only from a second aggregation level, but also from a not input enabled (advanced)  DSO. While reading reference data through an ’external’ aggregation level has been PAK enabled since BW 7.50 SP3 and BW 7.4 SP15, reading  not input ready DSOs still used the ABAP runtime. With BPC/4, reading reference data from not planning enabled advanced DSOs can now also be processed in HANA. Of course, planning enabled aDSOs should still be consumed by using Aggregation Levels.

The Fox coding is identical as with the ABAP case:

DATA PROD TYPE 0IP_PROD.

DATA SEGMNT TYPE 0IP_SEGMNT.

INFOPROVIDER NOT_PLANNING_DSO.

SEGMNT = VARV( SEG ).

FOREACH PROD.

{ PRICE, PROD } = NOT_PLANNING_DSO.{ PRICE, SEGMNT, PROD }.

ENDFOR.

 

Advanced DSO Planning Improvements and INT 8 Support

 

We support planning on advanced DSOs already since BW 7.5 SP4 (Cube-like and Direct Update aDSO). Details are described in note 2189829. Also InfoObjects with high cardinality, InfoObjects based on HANA views, and and also pure fields are allowed in planning providers with this SP.

 

For advanced DSO of type direct update also key figures with standard aggregation NO2 (e.g. for prices) can be used.  They are also leveraged as so called SID-based Key Figures for planning on attributes and short comments.

 

In such cases we now use a key-based planning buffer.  The buffer contains only characteristic and key figure values. The SIDs and navigational attributes are computed on demand later. With this new approach we gain the flexibility to support new scenarios for the usage of InfoObjects with BW 7.5 SP 8 and BPC/4. We can now combine InfoObjects and fields in a planning enabled aDSO and the HANA optimized runtime now also supports InfoObjects with transient SIDs and fields.

 

As you know, no navigational attributes are present at aDSO level. They are introduced by CompositeProviders on top.  InfoObjects with navigational attributes based on HANA views in combination with NO2 and SID based key figures are also supported on aDSOs from 7.50 SP8 onward. Each of the features could only be used separately before hand.

First customer are now adopting advanced DSOs. We have collected a number of notes we recommend all customers to implement. The collection can be found in note 2429714.

 

 

We also support key figures of type type INT8 for advanced DSOs.

In Planning, they may occur

• in planning enabled providers


• in providers used as reference data

The feature is  also delivered with BW4 SP4/BPC/4 SP0 and SAPBW 7.52. (warum ‘also’?

Auch in BW 7.50 SP8?)

 

Improvements for Input Ready Queries and New Lines on any Level

 

For input read queries, we offer 4 main improvements with BPC/4 and BW 7.5 SP8 onwards.

• New lines on any level
  An input enabled query can now have free characteristics which are in the aggregation level but not in the drill down. Only units and 0INFOPROV have to be drilled down or specified in the restriction. This is including intermediate sums or hierarchy nodes.  The values in new lines will be disaggregated. The disaggregation is then done equally on master data or characteristic relationships within the fixed filter


• Read posted, disaggregate unposted
  For each characteristic in a query you can define the way how data is read – either all posted values should be shown or all ‘possible’ value – either from the master data or the characteristic relationships (see extended properties, ‘Access type for result values’). You up to now you could only decide either to work on all existing records (posted) or a possible records (unposted, combination proposal). However it can make sense to control this separately. With the new flag on the query header level one can now decide to read only posted values but then disaggregate to all combinations (according to the master data or characteristic relationships) on the other side. Thus, you can display all existing records in a query but when you enter data on a (sub-)total level disaggregate to all possible combinations in the selection.


• Self-Reference in case of empty data in the cells/key figure-> Reference to other key figure for data access
  At the beginning of a plan cycle the plan data is usually empty. Thus, no matter what dissagregation type is chosen for a given key figure, when new data is entered on an aggregated level the system cannot disaggregate the data as no valid target records are existing. As usually there is data available for other structure elements (for example the actual data of the previous year) it is now possible to specify this reference structure element for defining which data records should be created. Please note that the relative values of the reference column is not taken into account. When you enter data on an aggregated level in a previously empty plan column then the reference column is used to create empty plan records and the the data is disaggregated equally (no matter which disaggregation type is chosen for the plan column). Once plan data is finally available, the chosen disaggregation type will be used when disaggregating newly entered values (e.g. self reference).


• F4 for new line might be different to F4 for filter value selection
  Up to now there was just one setting that was driving the value help (F4) in the query filter, the variable screen, or the new lines. But especially in new lines you might want a different behavior for the value help – you want to see value that are not yet posted. In the value help for the query filter you might rather want to see only values with existing data. With the new setting one can now differentiate various F4 helps for variables and for new lines.

The main feature new lines on any level and the option to always disaggregate unposted even when reading posted data are planning specific settings in the general query settings tab.



The input enabled key figure now has a new field (on the planning section) to specify a data reference in case the data on the key figure is empty. This is very useful in case of self reference when only data in other key figures (like actuals) is available. The data is then distributed equally.


The new setting for the value help in the new line is available is the settings for the characteristic (extended section). You can choose between leveraging master data or characteristic relationships independently from the chosen selection for filter values. Check the functional updates of the upcoming AO release for consuming this feature.



For details see also note 2447006.

 

Characteristic Relationship Maintenance in the BW Modelling Tools

 

We started the journey to have one design tool together with  BW/4HANA using the BW Modelling Tools. The query designer is already available in the BW Modelling Tools and we enabled the maintenance of the aggregation levels in BW 7.50 SP4. With BPC/4 onwards we will disable the creation of aggregation levels in RSPLAN.

With BPC/4 we now also have added the maintenance of characteristic relationships to the BW Modelling Tools. It is an own editor in the Modelling Tools which is opened through the context menu of the underlying advanced data store object. The search also first looks for the InfoProvider which then again can open the characteristic relationships.



 

All four types of characteristic relationships (Attribute-, aDSO-, Hierarchy or Exit based) can be maintained.

 

 

 

 

 

 

Improved Current Member Support

 

With BW 7.4 we introduced the concept of current member (as known from MDX) also in BW queries through a variable with replacement path. It was now possible to sum up values to the current period.

See for detail the SAP standard documentation or for planning the nice blog on SDN.

 

But one problem was open. Very often one only wants to sum up from beginning of the corresponding year or corresponding month. To overcome this problem, we have indroduced additional operators in the replacement rule for the first member of period ‘First Value’ and for the last member ‘Last Value’.

 

Let us look at an example where we report data from the beginning of the period e.g of the year to the current day (year to date). To achieve this, first one creates a variable of type replacement path with current member (this has been possible since BW 7.4 already)



The operator for the replacement rule is kept emtpy in our case.



Now we create a second similar variable with replacement path ‘Current Member’ . But we change the details on replacement and  change the operator to ‘First Value’.



Now we are ready to create our key figure and restrict it to the  days in the  allowed range from the beginning of the period (our second variable) to the current member (our first variable).



Now we can start the YTD reporting and look at the results here with fiscal periods and instead of days.

 

Input Ready Dynamic Restrictions (created by End User)

With BW 7.5 SP4 and Analysis Office 2.4 we started our roadmap for sheet based report definition which was described in the blog ‘Query less Reporting in BPC embedded’. The main part of your first delivery was the ‘restriction on the fly’ which is described in the blog  ‘Planning Engine Improvements for BPC Embedded Model, PAK and BW-IP with BW 7.50 SP4'.

So far it was only possible to create a ‘read-only’ restriction. Planning was not possible. With BW/4HANA SP4, BPC/4 and BW 7.5 SP7 it is also possible to create restrictions which are ready for input. However, those restrictions are only input ready if the underlying base key figure is also input ready. All disaggregation settings are inherited from the base key figure.

 

Dynamic Grouping

Beside the dynamic restrictions and the new AO feature of axis sharing, we have developed also the dynamic grouping in order to facilitate the sheet based definition of reports. In a similar way as with the MS Excel feature of grouping, you can now also bundle rows in the result set. By doing so, one is generating BW hierarchies on the fly which exist lives in the backend session of the BW or BW4/HANA system. This hierarchy is stored with the query view of the workbook and is recreated in each session. It behaves like a modeled BW hierarchy but is not visible in the BW hierarchy maintenance. Watch the movie on dynamic grouping if you want to get an impression. By the way grouping is also possible for objects which are not hierarchy enabled in BW. Using drag and drop in excel you can use this to build up the hierarchy.