Chars in a dimension

Hi guys,
  i have many chars in this cube as separate dimension, now my query(all chars in it 15) is taking lot of time to execute mostly for table scans done everything else, will it help if i club say brand and category in one dimension, wat are the pros and cons, wat things to keep in mind.
Thanks,
Your help is greatly appreciated

Hi,
Coming to the query performance you should take the following measures while creating a cube
1. No of dimensions should be less, which will reduce no. of joints in your reports.
2. No of records in each dimension should be less, which improves your database performance.
3. If only one character is asigned to a dimension and cordinality is 20% to that of the fact table we can go for line item dimensions, which directly assigns your SID to fact table without DID.
4. When the relation between the master table is 1-1 or 1-many, then go for seperate dimensions.
5. When the relation between the tables is Many - 1, then go for same dimensions.
Regards,
Pushpa M

Similar Messages

  • How to identify Master data chars in a cube?

    How to identify Master data chars in a cube? Is there any function module with teh help of which we can identify the master data objects in a cube. (I need to avoid manual checking as the no. of objects is more).
    Regards,
    R.Ravi

    right click on cube..click on 'display data model'..
    there expand the tree for each dimension..
    under it..u will see list of char in that dimension..
    for some characteristics u will see an arrow beside it..expand it..u will see further list of char..(with yellow icons)..
    these are the list of attributes for that particular characteristic..
    for char that have no attr..there will be no arrow beside it..
    by this..u can know which char in ur cube hold master data..
    Vishvesh

  • Regarding Line item dimensions

    Hi all,
    In our project for one of the cube 3 line item dimensions are maintained already , But still we have to create one more line item dimension for that infocube because the percentage for the one of the  dimension and fact table is reached to 90%.So is it preferable to have more  line item dimensions? If not please provide the alternative suggestion.
    we maintained relative infoobects in dimensions.
    Thanks & Regards,
    Swarna.P

    Hi,
    If there is no other alternative to swap the Chars in between Dimensions, then go for Line Item Dimension. It won't be a problem if you have more then 2 or 3 Line Item Dimensions. I Used 4 Line Item Diemsions becasue the Ratio is crossing 90%. So Line Item is solution for my problem, first check the Chars in Dimensions and re-arrange it then check the Ratio, if it doesn't work then go for Line Item Dimension.
    Thanks
    Reddy

  • Not able to merge dimensions objects on webi report based on SAP query

    Hello
    We are using SAP BI 7.0 , BO XI 3.1 sp1
    I have two queries in webi report. These two queries are based on the same universe. I am not able to merge dimensions from both queries as merge dimensions functionality is greyed. Two dimensions are of the same type (char) and are dimensions (not the details).
    Does anyone know the reason?
    Thanks

    Hi,
    First check your user ID has rights for merge dimension.may be you can try on other universe and see if it is still grayed out then possible rights issue.
    or may be you can check below thread.
    http://www.forumtopics.com/busobj/viewtopic.php?p=784845&sid=73804295827b0854dc3a654cf74fda92
    Thanks,
    Amit

  • Why do we put similar infoobjects into single dimension.

    Why do we put similar infoobjects(Char ex: 0material, 0matl_type, 0matl_group, 0matl_cat,  0mat_plant) into single dimension(MATERIAL DIMENSION). and other similar like(0CUSTOMER, 0SHIP_STCK, 0SOLD_TO, 0DIVISION, ZORD_MAT) into single dimension(CUSTOMER DIMENSION), What will happen if i put all these infoobjects into single Dimension, Can some one answer me in the system prospect, data base prospect  as well as performace point of view or may in general.
    Points will be given.............. thats what u definitely deserve for a right answer.

    Santosh,
    This is an important thing while deciding upon the dimensions. Its not the similar things we are putting together but based on the data realtion and cardinality we put the chars in a dimension. If dimensions are not properly designed the tables size increases drastically and lead to poor performance. Taking an example the relation between the chars in a dimension should be 1:m and not n:m. A material can have one material type and belons to one materila group for a plant is the relation which is a 1:n relation. Chars with high cardinality are put in separate dimensions instead of grouping with other chars.
    The grouping of classification groups (characteristics) whose content is logically connected, under a single generic term. If the dimension contains a characteristic whose value already uniquely determines the values of all other characteristics from a business-orientated viewpoint, then the dimension is named after this characteristic.

  • Check dimensions consistency in infocube

    Dear all,
    I need to check dimensions consistency in infocube.
    I know there're a lot of test in RSRV transaction, but I can't find which is right for me and what's the meaning of test (all are green).
    I need to check if I put right characteristics together (none relation NxM in dimension), because I'm implementing financial flow (FM and GM) and standard cubes (FM and GM) are discrepant (some characteristics alone in FM cube, the same together in GM).
    Thank you,
    Alessandra

    Hi,
    As per I understand you want to check how grouping the various char in the dimensions of the cube will affect the performace of the cube.... or what is the better way to group the infoobjects in the dimension.
    You cannot check this in program SAP_INFOCUBE_DESIGNS in SE38.
    Just execute it and search for your cube.
    it will give you the information about the data stored in all the dimension and the there percentage as related to fact table.
    Here the one which exceeds more then 20% will show up in red and you need to fix those dimensions and rearrange the char in them and shift to another dimension.
    this can be done only if the cube contains data and you should try it in production.
    Without loading cube you cannot decide right combination or you will have to reply upon your experience.
    Thanks
    Ajeet

  • Removing line item dimension

    Dear All,
    I have encountered a problem after I removed line item dimension flag from the dimension properties and added new characteristics to that dimension.
    When I checked the data in the cube, I was surprising to find the old characteristic (that was the only char. in the dimension previously) having irrelevant values.
    Now I can understand that there was no dimension table before removing the flag. And the new dimension table leaded to wrong matching between fact table and SID table.
    Can you suggest any solution to make the correct matching between the fact table and SID table ?
    thanks in advance..
    regards,
    berna

    Try performing the RSRV Tests.
    Go to tcode: RSRV and perform the following tests:
    All Combined Tests --> Transaction Data --> Foreign Key Relationship Between Dimension(s) and SID Tables for the Associated Characteristics
    All Combined Tests --> Transaction Data --> Fact and Dimension Tables of an InfoCube
    Cheers,
    Neel.

  • Dimension, Line item dimension

    Hi Friends,
    What exactly the line item dimension? when it is used? can we assign more than one char to a dimension? if so what would be the result? in what situations we do like this? please help me
    regards....

    Line item dimension refers to single field. This is nothing but the SID of the field get stored in the fact table , and there will no extra dimension , hence improves performance while reporting as select need not go thru extra dimension table..
    Line item dimension is used for those fields where the number of records is expected to be more..
    You cant assign more than one characteristic to line item dimension.. But you can assign more than in normal dimension..
    Hope this helps.
    Regards
    Manga(Assign points)

  • SAP_INFOCUBE_DESIGNS

    I have run the report SAP_INFOCUBE_DESIGNS.  An example of the results for my cube is shown below:
    SMMCIM1            /BIC/DSMMCIM11      rows:          9    ratio:          7  %
    SMMCIM1            /BIC/DSMMCIM12      rows:         57    ratio:         47  %
    SMMCIM1            /BIC/DSMMCIM14      rows:          4    ratio:          3  %
    SMMCIM1            /BIC/DSMMCIM1P      rows:          3    ratio:          2  %
    SMMCIM1            /BIC/DSMMCIM1T      rows:          2    ratio:          2  %
    SMMCIM1            /BIC/DSMMCIM1U      rows:          5    ratio:          4  %
    SMMCIM1            /BIC/ESMMCIM1       rows:          0    ratio:          0  %
    SMMCIM1            /BIC/FSMMCIM1       rows:        122    ratio:        100  %
    My cube consists of the following dimension tables:
    SMMCIM11     Location
    SMMCIM12     Material
    SMMCIM13     Batch
    SMMCIM14     Stock type
    SMMCIM15     Vendor
    SMMCIM16     Sold To
    My questions are as follows:
    1) It seems that if the dimension table is marked as Line Item it is not shown in this report.  Is there a report otherwise that I can run which can tell me whether I have correctly set Line Item dimension for these dimensions, if not what are the consequences if I have set this wrong
    2) In the example above the dimension below should be set to Line Item
    SMMCIM1            /BIC/DSMMCIM12      rows:         57    ratio:         47  %
    3) How can I determine for the line items whether it should also have a high cardinality setting
    Thanks

    Hi Niten,
    I would rather advice you to relook at the design than go in for "line-item" dimension. You can go in for line-item dimesnion only if you feel the design is correct. The disadvantage of a line-tem dim is that you can have only 1 char in the dimension. High cardinality would use the B-Tree indexes instead of the normal Bitmap indexes. High cardinality is advisable to use if you are using line item dim.
    Bye
    Dinesh

  • Any tips at this Reporting requirement gathering stage?

    Hello Experts,
    I have 10 different sources of data which will mostly come to me in flat files.
    Some of the data coming to me for <b>UPDATE</b> will be
    1. daily,           (3 of the sources)
    2. weekly,       (2 of the sources) 
    3. monthly      (2 of the sources) 
    4. quarterly.   (2 of the sources)
    Some of the reporting requirements are <b>(GENERATE REPORTS)</b>:
    a) As needed i.e. on demand by managers
    b) Monthly
    c) Quarterly
    I am still in the report requirement gathering phase and I would like some hints here:
    Q1. What are the Dos and Don’ts?
    Q2. Does it matter whether the various sources are loaded into Cubes or ODSes? Which is the better choices and for what reasons?
    Q3. Any special guide on the frequency of Updates?
    Q4. Can the data elements from the 10 sources be all placed in say, 2 or 3 dimensions and within a single Cube?  Any advantages of this?
    Q5. Can each of the data elements from the 10 sources be placed in say, 10 different ODSes? Any advantages of this?
    Thanks in Advance.

    Hi Amanda,
    Actually, I’d split the process into those stages:
    1.     gathering report forms that users want to have and a frequency of report generation.
    2.     collection information on available sources of data
    3.     analysis: do the sources will provide all data for forms required – simultaneously the list of reports that would be possible to create is made. Here you can take into account not only availability of resources, but also budget/time/work force limitation. It’ll determine the scope of implementation.
    4.     Conceptual model (with parallel excersises in BW).
    •     If some reports (and data in them) are completely different, don’t having common information, then consider creating separate data providers. 
    •     If some reports (and data in them) have something in common, then consider creating separate data providers with a multiprovider on them.
    •     If all data are linked to some key(s) (for example a document number), then consider creating a single cube
    •     Determine time granularity (are your weekly, monthly and quarterly data may be derived from the daily data? If yes, maybe these data (except daily) are redundant. If not, then probably these data supply some additional information to daily? – consider merging this additional info into daily cube (or probably through separate ODS or cubes).
    So, it’s a somewhat an answer to Q2 and Q5.
    Q3. You may load data just before (say a several days) the reports generation.
    Q4. See above about single cube. Additionally, the rule of thumb is to have more little dimensions than few huge dimensions. Consider spreading out your chars into different dimensions. If you’ll be able to have all data in a single cube, it would give you an enormous flexibility in building reports (especially, not mandatory ones, but ad-hoc reports on demand).
    Also you can visit www.saptips.com site. I downloaded “Tips on Implementing a BW-Based Management Reporting Solution (and Getting Your Users to Use It!)” but cannot find it in inet. If you want I can send it to you along with “SAP BW projects: dos and donts” (ppt-demo and transcript).
    Best regards,
    Eugene

  • What are the key steps & order to follow: changes in my flat file structure

    HI,
    I have a Cube which sits on ODS.
    In the ODS,  the are 6 characterisics: Char1, Char2,...., char6; and 3 key figures: kf1, kf2, and kf3.
    The ODS is loaded through a flat file.
    Through an update rule and startup routine, the ODS updates the cube.
    Now, I have a new requirement to add 2 new characteristics (Cha10, char20) and one 2 key figures (KF55, KF66)).i.e. the flat file will now be coming in with these new fields.
    I have an idea but this this is the first time I really have to implement, I need to be sure.
    What are the key steps that I need to go through and in what order?
    Thanks

    What version are you running ?
    Will you need to load history data as for these new fields or is it just going forward ?
    Is it all one to one mapping for the new fields ?
    1. Make sure you know to what dimensions you need to add the new chars ? New dimension for these two ?
    2. New chars to be made keyfields ?
    3. Update Type for keyfigures Overwrite or Additive ?
    Enhance the Cube, DSO, change  TRFN/TR/UR..

  • Issue with Compression of Cube 0IC_C03

    Dear Mates,
    Before initiating this Thread, i tried to search on SDN but couldnot get something that will work in my case.
    Issue: Compression scheduled in Process Chain starts, runs for a long time and then gets cancelled,as i can see in Process (right Click) >> Display messages.
    Compression was running fine since last one year. It started with the above mentioned behavior since Nov 2010.
    One Area which i feel fishy, is in November we had a Database movement to DB2. But i really couldnot relate the Issue with compression to Change in Database.
    We are currently working on SAP BW 3.5
    Please share your Comments/ Suggestion. Below is the Log observed after Compression cancelled.
    Job log     Job log     Message text uncoded
    12/19/2010     03:03:48     Job started
    12/19/2010     03:03:48     Step 001 started (program RSPROCESS, variant &0000000530859, user ID BW_CPIC)
    12/19/2010     03:03:48     Performing check and potential update for status control table
    12/19/2010     03:03:58     FB RSM1_CHECK_DM_GOT_REQUEST called from PRG RSSM_PROCESS_COMPRESS; row 000200
    12/19/2010     03:03:58     Request '758.334'; DTA '0IC_C03'; action 'C'; with dialog 'X'
    12/19/2010     03:03:58     Leave RSM1_CHECK_DM_GOT_REQUEST in row 70; Req_State ''
    12/19/2010     03:03:58     FB RSM1_CHECK_DM_GOT_REQUEST called from PRG RSSM_PROCESS_COMPRESS; row 000200
    12/19/2010     03:03:58     Request '758.348'; DTA '0IC_C03'; action 'C'; with dialog 'X'
    12/19/2010     03:03:58     Leave RSM1_CHECK_DM_GOT_REQUEST in row 70; Req_State ''
    12/19/2010     03:03:58     FB RSM1_CHECK_DM_GOT_REQUEST called from PRG RSSM_PROCESS_COMPRESS; row 000200
    12/19/2010     03:03:58     Request '761.202'; DTA '0IC_C03'; action 'C'; with dialog 'X'
    12/19/2010     03:03:58     Leave RSM1_CHECK_DM_GOT_REQUEST in row 70; Req_State ''
    12/19/2010     03:03:58     FB RSM1_CHECK_DM_GOT_REQUEST called from PRG RSSM_PROCESS_COMPRESS; row 000200
    12/19/2010     03:03:58     Request '763.019'; DTA '0IC_C03'; action 'C'; with dialog 'X'
    12/19/2010     03:03:58     Leave RSM1_CHECK_DM_GOT_REQUEST in row 70; Req_State ''
    12/19/2010     03:03:58     FB RSM1_CHECK_DM_GOT_REQUEST called from PRG RSSM_PROCESS_COMPRESS; row 000200
    12/19/2010     03:03:58     Request '763.397'; DTA '0IC_C03'; action 'C'; with dialog 'X'
    12/19/2010     03:03:58     Leave RSM1_CHECK_DM_GOT_REQUEST in row 70; Req_State ''
    12/19/2010     03:04:06     SQL: 19.12.2010 03:04:06 BW_CPIC
    12/19/2010     03:04:06     INSERT INTO "/BI0/L0IC_C03" ( "SID_0REQUID"
    12/19/2010     03:04:06     ,"SID_0PLANT" ,"SID_0CALDAY_F" ,"SID_0CALDAY_T" )
    12/19/2010     03:04:06     SELECT 2000000000 AS "SID_0REQUID"
    12/19/2010     03:04:06     ,"/BI0/L0IC_C03"."SID_0PLANT" , MIN (
    12/19/2010     03:04:06     "/BI0/L0IC_C03"."SID_0CALDAY_F"  )  AS
    12/19/2010     03:04:06     "SID_0CALDAY_F" , MAX (
    12/19/2010     03:04:06     "/BI0/L0IC_C03"."SID_0CALDAY_T"  )  AS
    12/19/2010     03:04:06     "SID_0CALDAY_T" FROM "/BI0/L0IC_C03" WHERE (
    12/19/2010     03:04:06     "/BI0/L0IC_C03"."SID_0REQUID" BETWEEN 0 AND
    12/19/2010     03:04:06     763397 ) GROUP BY "/BI0/L0IC_C03"."SID_0PLANT"
    12/19/2010     03:04:06     SQL-END: 19.12.2010 03:04:06 00:00:00
    12/19/2010     03:04:06     SQL: 19.12.2010 03:04:06 BW_CPIC
    12/19/2010     03:04:06     INSERT INTO "/BI0/L0IC_C03" ( "SID_0REQUID"
    12/19/2010     03:04:06     ,"SID_0PLANT" ,"SID_0CALDAY_F" ,"SID_0CALDAY_T" )
    12/19/2010     03:04:06     SELECT -1 AS "SID_0REQUID"
    12/19/2010     03:04:06     ,"/BI0/L0IC_C03"."SID_0PLANT" , MIN (
    12/19/2010     03:04:06     "/BI0/L0IC_C03"."SID_0CALDAY_F"  )  AS
    12/19/2010     03:04:06     "SID_0CALDAY_F" , MAX (
    12/19/2010     03:04:06     "/BI0/L0IC_C03"."SID_0CALDAY_T"  )  AS
    12/19/2010     03:04:06     "SID_0CALDAY_T" FROM "/BI0/L0IC_C03" WHERE (
    12/19/2010     03:04:06     "/BI0/L0IC_C03"."SID_0REQUID" BETWEEN 0 AND
    12/19/2010     03:04:06     763397 ) GROUP BY "/BI0/L0IC_C03"."SID_0PLANT"
    12/19/2010     03:04:06     SQL-END: 19.12.2010 03:04:06 00:00:00
    12/19/2010     03:04:07     SQL: 19.12.2010 03:04:07 BW_CPIC
    12/19/2010     03:04:07     TRUNCATE TABLE "/BI0/0100000095"
    12/19/2010     03:04:07     SQL-END: 19.12.2010 03:04:07 00:00:00
    12/19/2010     03:04:12     SQL: 19.12.2010 03:04:12 BW_CPIC
    12/19/2010     03:04:12     TRUNCATE TABLE "/BI0/0100000091"
    12/19/2010     03:04:12     SQL-END: 19.12.2010 03:04:12 00:00:00
    Thanks & Regards
    Sameer
    Edited by: Sameer A Ganeshe on Dec 28, 2010 10:51 AM

    Hi Zeeshan,
    I  handled inventory scenarios, actually total stock qty is non *** Key figure with cumulative key figures of Issue qty and receipt qty, receipt - issues gives you the the total stock qty, and in inventory cube you can capture daily movments as well as monthly since you do have the calmonth as Time char in the dimensions.
    Regards
    Ram

  • Line Item reporting, how?

    Hi ...
    My cube has sales order, item in addition to other 50 charecterstics.
    1. Sales Order
    2. Sales Ordr item.
    I have made enough line items including sales order number but sales order item is causing the fact table grow.
    I can not make sales order item as line item (or) inlcude in sales order dimension.
    could you guys give me some advice which dimension i can put the sales order item field?
    thanks in advance.

    Hello,
    Sounds like you have the following setup:
    You already have sales order number as a line item, thus the only characteristic in the dimension it resides in (can only have 1 char in a dimension identified as line item).  You also have sales order item in your cube, but not as line item and it's causing your cube to grow to big.  Yes, having the sales order number is growth itself, but then to have the item number in the cube will cause huge growth because you can have many, many items per order.  If you cannot make line item for this char the best option is taking item number out of the cube. 
    If you load to a DSO and then push to a cube, leave the item number in the DSO and only push sales order number and other char to the cube.  Now, when a report is needed you can report to report interface to the DSO to get the line item information.  Meaning reports written off the cube give the user a particular report output.  Then, if they need line item, then can perform RRI to go to the DSO to get line item information.  Do you know how many reports require line item number?  Percentage wise?  If the majority of your reports require line item right away then you'll have to re-design and make a line item dimension.  If the percentage of reports is small that need line item then leave it in the DSO and take out of the cube.
    Hope this help!

  • What is the fact table content?

    a.     Key figure for a combination of char value of dimensions are stored in the fact table
    b.     Both cumulatiove and also key figure for non cumulative values can be contained in fact table
    Choose any one
    Thanks
    Babu

    Hi,
    Fact table contains the data (key figs) for a certain combination of characteristic values of the dimension tbls to help a business person evaluate their company & make appropriate decisions.
    => a. from the choice you had.
    Rgds,
    Colum
    Was this information helpful to you Babu?
    If so, please award points as appropriate.
    thanks
    Edited by: Colum Cronin on Apr 7, 2010 10:11 AM

  • Not getting the EDIT INFOCUBE screen

    Hi All,
    Im new to biw. I m referring to the Step by Step guide for BIW book. There are lots of difference in the screens since i use ECC 6.0. I have created an InfoArea, Info Object Catalog and Info Objects. now when i create the InfoCube, after creation it does not go to the EDIT InfoCube Screen. I could not move any further in the document.
    Please help me out.
    Regards
    RS

    Hi,
    This is means that u have not added any key figures and characteristiccs to that cube and u have not assigned any chanractersitic to that dimension... Go to the characteristic tab and drag the characteristics and KF's and assign the char to the Dimension and activate the Cube....
    Assign points if it helps
    Khaja

Maybe you are looking for