Using K Means in a Calculation View

Hi Experts,
Can you please help me with the below scenario.
I am trying to use K-Means function to perform clustering on customer data. I have the below columns:
Attributes:
Customer ID, Home Ownership Status
Measures:
Customers Debt to Income Ratio, Annual Income
1. Is it possible to use multiple attributes given above to perform the clustering as i read that clustering can be only done using measures.
2. I need to pass the number of clusters dynamically from report to the K-Means function to perform clustering. I tried to use calculation view to get this parameter from user at runtime and pass to the input parameter table of K-Means function. But the problem is, calculation views will not allow any kind of DDL statements like create/drop/insert/delete/update etc.
3. Further i would like to understand how to use the results of the K-Means clustering delivered via two tables one having clusters, distance and other having clusters and remaining measures to build a scatter chart. How this can be achieved.
Please help.
Thanks and Regards,
M.N.Adinarayanan

Hello,
Would you share your dataset maybe I can help.
Kind regards.

Similar Messages

Errow with IF statement in Scripted Calculation view

Hi Team,
I am trying to use IF statement in Scripted Calculation view.
My scenario is:-
I need to create a restricted Key figure based on input date from user.
While i am activating my view system gives "Internal deployment of object failed;Repository: Encountered an error in repository runtime extension;Internal Error:Deploy Calculation View: SQL: transaction rolled back by an internal error: Syntax Error for calculated Attribute: line 29 col 9 (at pos 1501)" error.
Below is the code i have used in my CV.
CE_CALC('IF((to_int("BUDAT") >= to_int(2013-01-01),"AMOUNT",'''')', DECIMAL(13,2)) AS "DTR"
Here, "AMOUNT" is another calculated field created using CE_CALC.
Without above IF statement my CV runs perfectly fine. The moment i include IF statement it does not activate.
Kindly provide your inputs.
Thanks & Regards,
Anup

Hi Lars,
I am extremely sorry about the confusion.
There was an extra parenthesis in the IF statement which was not allowing the CV to be activated. CV is now successfully activated.
But my data is not getting restricted correctly.
It is showing 0 against all the records.
Below is the sql used for calling the CV.
SELECT * FROM "_SYS_BIC"."A_123/CV_SCR_BSID_BSAD_ITEMS" ('PLACEHOLDER' = ('$$Currency$$', '''Local'''),'PLACEHOLDER' = ('$$Date_From$$','2014-01-01' ), 'PLACEHOLDER' = ('$$Date_To$$','2014-06-30' ));
Below is my updated CE_CALC function :-
CE_CALC('IF(date("BUDAT") >= date(:Date_From) and date("BUDAT") <= date(:Date_To),"AMOUNT",0)', DECIMAL(13,2)) AS "DTR"
Is there any issue while passing values here?
Regards,
Anup

Check Column Type in Calculation View

Hi All,
Before coming to actual problem, let me tell you a little background about the scenario.
Tables are replicated using SLT from a Java based system (SAP MII) to HANA. The table (say TABLE_1) contains a column of type DATE.
SLT be default converts the DATE type to NVARCHAR while replicating.Hence in HANA the column type is NVARCHAR.
Now, in HANA I am converting this column to SAP_DATE format (YYYYMMDD) by using the formula
leftstr(replace("COLUMN_NAME",'-',''),8)
Problem:
The MII (The source system) can also be on HANA. So there are 2 scenario possible.
MII is using some other DB and it is connected to HANA using SLT.
MII is itself is on HANA and no SLT is required.
In scenario 2, since there are no SLT, the column type is DATE in HANA. Therefor the expression leftstr(replace("COLUMN_NAME",'-',''),8) gets failed.
We need to check the data type of column and then change the expression dynamically. For example
If (COLUMN-DATA-TYPE = DATE)
convert to string and the apply the expression leftstr(replace("COLUMN_NAME",'-',''),8)
Else If (COLUMN-DATA-TYPE = NVARCHAR)
apply the expression leftstr(replace("COLUMN_NAME",'-',''),8)
There are system tables/views which can be used to check column data type. But all there are Row-based. We can not use row-based table in calculation view.
Also if we use them some additional privileges will be required for all end-users who need to check output of the views.
Awaiting suggestion/ideas. Thanks.
Regards,
Raja

Hi Raja,
TO_DATS is not available as an expression for graphical calc view.
It is a SQL function.
Question is: what do you need the legacy format for?
Why not use the date data type?
The problem with using the non-date data types to handle date information is that the database cannot know that it is a date column.
Validity checks don't work.
'20141301' is completely fine with the database if it is stored as string - but not as a date.
Cardinality estimations will be off.
How many distinct values can be expected between '20140101' and '20140201' ?
The date data type "knows" there are only 30 further values possible.
The string representation is open for at least a hundred values - if you just stick to the numeric digits.
Also, when you use this column in your graphical calc view, the frontend UI won't know that it is supposed to be date column.
No fancy date picker widget will pop up.
Finally, this is slowing down your model.
The conversions need to be done over and over - what for?
All in all, yes, I do see some problems with this approach...
So, you really should consider to perform the conversion before you store the data.
Details on how to configure SLT can be found in, you guessed it, the documentation!
- Lars

Error while creating a BO Universe (using IDT) on top of SAP HANA Calculation View with Input Parameters

Hi All..
We are trying to create a Universe (using IDT-Version4.0) on top of the HANA Calculation view. The Calculation view has 4 input parameters, So
in Universe side we have created the respective prompts. For the creation of derived table we have used the following code
SELECT *
FROM "_SYS_BIC"."<schema_name>.<CV_Calculation_View_test>"
'PLACEHOLDER'=('$$IP_A$$','@Prompt(A)'),
'PLACEHOLDER'=('$$IP_B$$','@Prompt(B)'),
'PLACEHOLDER'=('$$IP_C$$','@Prompt(C)'),
'PLACEHOLDER'=('$$IP_D$$','@Prompt(D)')
While validating the above code we are getting an error.
I have attached the snapshot of that error for your reference. Please find the attachment and help me in resolving it.
Thanks in advance.

Hello George,
I don't have any personalization set on the info space. Also I am using mapped Account in BI to connect to HANA. The confusing part is that it is able to validate the infospace with the input parameters and index it. However query does not return any results. I even tried running the same query which explorer sends to HANA in the SQL editor and there too the same results,the query does not return anything. The model does return data when I do a data preview and if accessed from other tools like AAO.
Also when I use SSO connection to HANA, indexing of the infospace fails. Where can I see the error log?
Thanks,

Calculation View computing 1 million net present values using BFL procedure

Hi,
i have around 1 Million financial positions in a position table and for each of them i want to calculate the net present value using SAP HANA BFL procedure _SYS_AFL.AFLBFL_NETPRESENTVALUE_PROC. The necessary cash flows, interests, etc. (used as input for the BFL procedure reside in a cash flow table
To make it nice i would like to use a HANA calculation view, which provides the position table together with a calculated field NETPRESENTVALUE
Now my question: How can i do this the fastest way? The challenge i see is, that the BFL procedure allows only the calculation of one net present value result at a time (based on the cash flows of one financial position).
So i am afraid i have to use
DECLARE CURSOR l_position_cursor ...
    FOR l_position_row as l_position_cursor DO
         CALL _SYS_AFL.AFLBFL_NETPRESENTVALUE_PROC ...;
    END FOR;
Do you have any better / more performant ideas?
Thanks for any idea,
Gunther

Does the ext directory have the php_oci8.dll? In the original steps the PHP dir is renamed. In the given php.in the extension_dir looks like it has been updated correctly. Since PHP distributes php_oci8.dll by default I reckon there would be a very good chance that the problem was somewhere else. Since this is an old thread I don't think we'll get much value from speculation.
-- cj

Using Hierarchy in Calculation View

Hi guys,
Can any one let me know the importance of Hierachy in the Calculation View.
I have created a Hierachy with the Hierachy type as Level Hierarchy and have set the elements manually in it.
I dont see any changes in the resultant data. Preview of data is as such similar to that when i was not using the Hierachy level.
Can any one guide me in this!!
Thanks and regards
Sanjay

You wil not find any difference in the "Data Preview".
Tools like Excel or Analysis office will help you to see the difference
Regards,
Krishna Tangudu

Using Attribute views in Script based Calculation views

I am trying to create a view for analysis on a fact table which has measure name as a dimension as opposed to having multiple measures. Here is an example of what the data looks like.
#Customer, Product, Measure, Data
Customer1, Product1, ASP, 4.2
Customer1, Product1, Units, 100
Customer2, Product2, ASP, 4.4
Customer2, Product2, Units, 200
... and so on and so forth.
I had build Attribute views with hierarchies for both customer and product dimensions and build analytic views. However now I have the need to flatten the Measures dimension and have the resulting Fact view look like:
#Customer, Product, ASP, Units
Customer1, Product1, 4.2, 100
Customer2, Product2, 4.4, 200
... and so on and so forth.
Now I want to build analytic views on this flattened view of my Fact table.
I was able to create a calculation view using Script mode, on the fact table and write a SQL to flatten the Measure dimension. But I am not able to now join this calculation view with the attribute views. Is there a way to leverage the existing Attribute views and their hierarchies in this case?
The only way I can see for now is to use manual SQL to build out the entire view, including joining the attribute views and re-build the hierarchies in the definition of the calculation view. This essentially wastes work I have already done in the attribute views and forces me to build the entire thing using SQL.
Is there a better approach?

It's extremely annoying that the FLD() command isn't yet availlable in NW, unless it's coming with 7.5 or has been included in one of the more recent support packs.
As a an alternative you may be able to use the following MDX routine:
Returns the value of a property of the current member of a dimension
*function PRO(%DIMENSIONNAME%,%PROPERTYNAME%)
%DIMENSIONNAME%.Properties("%PROPERTYNAME%")
*endfunction
Unfortunately I've never been able to use this with a When statement, so you may have to ask an ABAP resource to develop a BADI to achieve your required results.

Scheduling AFO that uses HANA Calculation View

We are using Analysis for Office against a HANA Calculation view.
We are saving the Analysis for Office workbooks on the BI platform.
I am able to create a publication, schedule the workbook emailing it to myself. It shows that it successfully completed, but I never get an email.
I am able to open the workbook from the BI Platform in AFO and it does refresh.
Any ideas?

are you able to see the output of schedule in the History?
Do you have any issues with the SMTP? are you able to get the Email from BO server?
refer the below doc.
http://scn.sap.com/docs/DOC-44785
BI Publication on 4.1 running successful but not sending emails to outlook

Using calculation view in excel with input parameters

Hello
I have a calculation view which receives 2 input parameters for a range of dates, when i create the connecction with Excel i get the error "sql processing error: Hierarchy source query: SQL: column store error: search table error: [34092] search on calculation model requires parameters: Required variable $$MyVariable$$ is not set"
I would like to know how can i pass the parameteres from excel to my calculation view??
Thanks

I doubt this can be achieved with a native Excel connection. Try SAP Advanced Analysis for Office. (I believe it may now just be called Analysis for Office).

How to call a Calculation View from Sql ?

Hi All,
I'm trying to test my calulation views. Now i'd like to call it through sql. How to do it ?
Thanks
Sakthivek

Hi Sakthivek,
After you activate your calculation view successfully, you will find it under schema "_SYS_BIC" -> Column Views in your Studio. Then you can use select sql statement to query the calculation view.
Best regards,
Wenjun

Problem with using multiple Entity Objects in a view Object.

Hi
Thank you for reading my post
I have create 3 Business components for 3 of my database tables and now
I must add 3 tables in a View object so i used Jdeveloper Wizard to create the View Object.
-I Add Entity Objects which are business components to this view (In Step 2 of the Create Vview Object wizard).
-In step 3 that I add Attributes all my attributes are marked as Transient
Can some one explain why it happens?
I need one of those tables to be updateable and two other tables are not updateable.
What should should i do to achieve this?
I should say that tables does not have any database relation (Foreign Key I Mean).
Thanks.

Hi user505214
When you created your VO, on selecting the second EO, you'll note at the bottom of the same page on the wizard/editor, checkboxes for updatable or by reference. By default reference is checked and this will make your second EO's attributes transient.
In the JDeveloper Developer's Guide for 4GL/Forms programmers, the following sections outline the difference between updatable or by reference:
7.5 Including Reference Entities in Join View Objects
27.9 Creating a View Object with Multiple Updatable Entities
Make sure to read 27.9 if updatable is what you want as it indicates you may need to add some additional code.
Hope this helps.
CM.

Hierarchy on HANA Calculation View with Optional Input Parameters Fails

Hi,
Has anyone succeeded in building a hierarchy on top of a calculation view with optional input parameters, where an input parameter is not filled?
The original requirement came from the wish to create a parent child hierarchy on a calculation view that was copied into the customer space from a HANA Live Financial Statement query view, but I have found the following when creating a simple level hierarchy on a calculation view that consumes one table. The Calculation View has one Input Parameter where the 'mandatory' box is unchecked.
Calculation view reads ECC table FAGLFLEXT
Simple level based hierarchy on fields PRCTR, RACCT and SEGMENT
Input parameter is used as a filter for PRCTR with logic ("EMPTY" = '$$P_PROFITCTR$$' or "PRCTR" = '$$P_PROFITCTR$$')
When I run the view I see the following behaviour in HANA Studio and Analysis for Excel
Before building the hierarchy I could run the view with or without the Input Parameter
After building the hierarchy I can run the view with the Input Parameter filled, but it fails when the Input Parameter is not filled.
Error message is "error: search table error: [2426] missing placeholder; missing value for mandatory parameter P_PROFITCTR"
P_PROFITCTR is not a mandatory parameter, but the selection for the hierarchy view thinks it should be. I don't find any notes around this issue, so I don't think it's version related, however the version I have used for this test is 1.00.70.
It's a shame we can't currently build the hierarchy as the parent child relationship is provided in HANA Live view NewGLFinancialStatementQuery.
Thanks,
Ken

Hi Ken,
We have been facing similar issue. We have even tried to set default value and as optional parameter. View still fails to create hierarchy (we are not using HANA live). As mentioned on page 97 of HANA modelling guide input parameter is mandatory from engine point of view. Hierarchy is generated as column view during the initial activation of calculation view and therefore expect a value by caller.
This seems to be a product error. Some one from HANA development team should explain this issue in detail. I would expect someone like Thomas Jung reply to us.
At the moment, we cant use input parameter for date prompt which gives us calendar popup feature for date selection. Hierarchies just don't work with input parameter. We are missing something.
Regards
Angad

Out of memory Error while querying SQL Script based Calculation View

Hi All,
I wanted to test the performance of Graphical and SQL Script based Calculation views.
Created Graphical (CA_GRPH) and SQL Script (CA_SQL) Calculation views.
Analytic View (AN_GRPH) for both Calculation views are the same which is Graphical based (90 Attributes and 5 Measures)
In Analytic View data foundation I have a Fact table which has 1.5 Billion records and 9 Dimension Tables –collectively 500 million records (7 Attribute Views). 9 Referential joins with cardinality N:1 and 1 Referential join with cardinality N:N.
I wanted to keep (CA_GRPH) and (CA_SQL) as a base Calculation views and leverage those to create various calculation views (Will be creating different Calc views for respective Business segments)
In order to test this I have created below calc views on top of base calc views.
Graphical Based: Created (CA_GRAPH_XYZ) by having CA_GRPH in projection with 30 Calculated Columns. – This retrieves data in 13 secs
SQL Script Based: Created (CA_GRPH_ABC) by having CA_SQL in projection view with 30 calculated columns – This errors out after 1.50 mins.
Could not execute 'SELECT "COLUMN_A","COLUMN _B"," COLUMN _C"," COLUMN _D", SUM("COLUMN _REVENUE") AS ...' in 1:50.480 minutes .
SAP DBTech JDBC: [2048]: column store error: [2048] column store error: search table error: [1000002] Error executing physical plan: exception 1000002:
ltt/impl/memory.cpp:63
Out of memory ; $size$=1507711; $name$=ihm; $type$=pool; $inuse_count$=170104; $allocated_size$=219215007925
exception 1000002:
Any suggestion / help in fixing this issue will be greatly appreciated.
Regards,
Av

Hi Raj,
Thanks for your time, please find edited snap hot of Analytic View (AN_GRPH) below,
Calculation view(CA_SQL)
        /********* Begin Procedure Script ************/
BEGIN
       var_out =
     SELECT
            "COLUMN_1"
            "COLUMN_2",
            "COLUMN_84",
            "COLUMN_85;",
            SUM("REVN") AS "REVN",
            SUM("MGN") AS "MGN",
            SUM("ORD_QTY") AS "ORD_QTY",
            SUM("SYS_QTY1") AS "SYS_QTY1",
            SUM("SYS_QTY") AS "SYS_QTY"
FROM
      "_SYS_BIC"."XYZ/AN_GRPH"
GROUP BY
            "COLUMN_1"
            "COLUMN_2",
            "COLUMN_84",
            "COLUMN_85";
END
/********* End Procedure Script ************/
Later i have built one more Calculation view(CA_GRPH_ABC) using (CA_SQL)in projection. i have 30 calculated measures in this final calc view. this final calc view is throwing above mentioned error.
Not sure if i can use SQL script based calc view in graphical based calc views?
Regards,
AV

SAP HANA: Error while activating Graphical Calculation Views

Hello,
I am receiving the below error while activating a Graphical Calculation View.
I am getting this error lately. Not sure what is causing this because the calculation views worked just fine before.
It is failing at the Repository Activation step with the Activation ID 2146.
It says - Repository: Revalidation failed for at least one object.
MDXRuntime: Internal deployment of object failed: Internal Error:Create Scenario: failed aCalcEngine.createScenario(): The following errors occured: Could not create repository entry for scenario (34053)nnprinting XML <?xml version="1.0" encoding="UTF-8" standalone="no"?><cubeSchema defaultLanguage="EN" defaultSchema="_SYS_BIC" operation="createCalculationScenario" version="3"><calculationScenario name="sflight/GC_AGENCY_REVENUE" schema="_SYS_BIC"><dataSources><tableDataSource name="SBOOK" schema="SFLIGHT" table="SBOOK"><attributes><allAttribute /></attributes></tableDataSource><tableDataSource name="STRAVELAG" schema="SFLIGHT" table="STRAVELAG"><attributes><allAttribute /></attributes></tableDataSource></dataSources><calculationViews><join name="Join_1" type="inner"><viewAttributes><viewAttribute ignoreFromInputsFlag="false" isVisible="true" name="MANDT" /><viewAttribute ignoreFromInputsFlag="false" isVisible="true" name="AGENCYNUM" /><viewAttribute ignoreFromInputsFlag="false" isVisible="true" name="FORCURAM" /><viewAttribute ignoreFromInputsFlag="false" isVisible="true" name="LOCCURAM" /><viewAttribute ignoreFromInputsFlag="false" isVisible="true" name="NAME" /></viewAttributes><inputs><input name="SBOOK"><mapping source="MANDT" target="MANDT"/><mapping source="AGENCYNUM" target="AGENCYNUM"/><mapping source="FORCURAM" target="FORCURAM"/><mapping source="LOCCURAM" target="LOCCURAM"/></input><input name="STRAVELAG"><mapping source="MANDT" target="MANDT"/><mapping source="AGENCYNUM" target="AGENCYNUM"/><mapping source="NAME" target="NAME"/></input></inputs><joinAttributes><joinAttribute excludeFlag="false" name="MANDT" /><joinAttribute excludeFlag="false" name="AGENCYNUM" /></joinAttributes></join><aggregation name="finalAggregation" defaultViewFlag="true"><inputs><input name="Join_1"><mapping source="NAME" target="NAME"/><mapping source="AGENCYNUM" target="AGENCYNUM"/><mapping source="MANDT" target="MANDT"/><mapping source="LOCCURAM" target="LOCCURAM"/><mapping source="FORCURAM" target="FORCURAM"/></input></inputs><viewAttributes><viewAttribute name="NAME"/><viewAttribute name="AGENCYNUM"/><viewAttribute name="MANDT"/></viewAttributes><keyfigures><keyfigure name="LOCCURAM" aggregationType="sum"/><keyfigure name="FORCURAM" aggregationType="sum"/><keyfigure name="row.count" datatype="int" isVisible="false" aggregationType="count"/></keyfigures><calculationNodeHints countStarAttribute="row.count"/></aggregation></calculationViews></calculationScenario></cubeSchema>nVersion: 33n
I am using the SFLIGHT schema which is provided by SAP.
Any pointer is appreciated.

Hi Murali
I met the exactly same issue when I try to activate a graphical calculation views.
I created a new user in HANA db, then create a simple graphical calculation view(just from one table from this user). It doesn't work. All attribute views work fine. There's no related column view under _sys_bic.
HANA database version is 1.00.31 runs on HP DL580 512GB. And here's the error message -
"MDXRuntime: Internal deployment of object failed: Internal Error:Create Scenario: failed aCalcEngine.createScenario(): The following errors occured: Could not create repository entry for scenario (34053)nnprinting XML <?xml version="1.0" encoding="UTF-8" standalone="no"?><cubeSchema defaultLanguage="EN" defaultSchema="_SYS_BIC" operation="createCalculationScenario" version="3"><calculationScenario name="co-pc-river/CALC_TEST1" schema="_SYS_BIC"><dataSources><tableDataSource name="JEST" schema="GLR" table="JEST"><attributes><allAttribute /></attributes></tableDataSource></dataSources><calculationViews><projection name="finalProjection" defaultViewFlag="true"><inputs><input name="JEST"><mapping source="MANDT" target="MANDT"/><mapping source="OBJNR" target="OBJNR"/><mapping source="STAT" target="STAT"/><mapping source="INACT" target="INACT"/></input></inputs><viewAttributes><viewAttribute name="MANDT"/><viewAttribute name="OBJNR"/><viewAttribute name="STAT"/><viewAttribute name="INACT"/></viewAttributes></projection></calculationViews></calculationScenario></cubeSchema>nVersion: 64n"
Best Regards
Mao Mingqi

Getting Error while creating Calculation view.

Hi Expert,
I was trying to create one simple calculation view using SQL script but getting below error.
/********* Begin Procedure Script ************/
BEGIN
create type var_out1 as table (ebeln varchar(10),menge decimal);
        var_out = CE_COLUMN_TABLE ("SAPEH2"."EKKO",[EBELN,AEDAT] );
        var_out1 = CE_COLUMN_TABLE ("SAPEH2"."EKPO",[EBELN,MENGE] );
        var_out2 - CE_UNION_ALL (:var_out1,:var_out2);
END
/********* End Procedure Script ************/
I thought , i might need to declare a table type for output table var_out1 and var_out2. I tried by declaring table like below couldn't get any success.
create type var_out1 as table (ebeln varchar(10),menge decimal);
create type var_out2 as table (ebeln varchar(10), aedat DATE ,menge decimal);
error:
Message :
     Internal deployment of object failed;Repository: Encountered an error in repository runtime extension;Internal Error:Deploy Calculation View: SQL: sql syntax error: CREATE TYPE is not allowed in SQLScript: line 5 col 2 (at pos 277)nSet Schema DDL statement: set schema "SAPEH2"nType DDL: create type "_SYS_BIC"."pkg-sapuser03.test_giri/CA_SALES1/proc/tabletype/VAR_OUT" as table ("EBELN" VARCHAR(10), "AEDAT" DATE, "MENGE" DECIMAL(13,3))nProcedure DDL: create procedure "_SYS_BIC"."pkg-sapuser03.test_giri/CA_SALES1/proc" ( OUT var_out "_SYS_BIC"."pkg-sapuser03.test_giri/CA_SALES1/proc/tabletype/VAR_OUT" ) language sqlscript sql security definer reads sql data as n /********* Begin Procedure Script ***********/ n BEGIN n n create type var_out1 as table (ebeln varchar(10),menge decimal);n create type var_out2 as table (ebeln varchar(10), aedat DATE ,menge decimal);n n t   var_out = CE_COLUMN_TABLE ("SAPEH2"."EKKO",[EBELN,AEDAT] );n t   var_out1 = CE_COLUMN_TABLE ("SAPEH2"."EKPO",[EBELN,MENGE] );n t   var_out2 - CE_UNION_ALL (:var_out1,:var_out2);nnEND /******** End Procedure Script ************/n
Could you please help me to understand the cause of this?
Thanks,
Girdhari

Hi Lars,
Thanks for the reply.
If i understood correctly , you are suggesting to create var_out1 and var_out2 in output ? If yes can you please guide me on how to do that? I tried but did not find anything suitable to declare table.
Thanks,
Girdhari

Using K Means in a Calculation View

Similar Messages

Maybe you are looking for