Ago function - Aggrgate Table.

Hi , tnks in advance for your replies.
I have one aggregate table(created in by db oracle 10g/r2) where i have the sum of "qty" and "sale_price" in a level of month.
I have a calendar dimension --->year->half->quarter->month(where month is my lower level) in my business level.
i created the appropriate logical columns with the ago function for each(year ago....) in administrator tool
Everything is ok for the "year-half-quarter" but when i try to create a report with my lowest dimension "month" , it is created the following nqserror :
Κατάσταση: HY000. Κωδικός: 10058. [NQODBC] [SQL_STATE: HY000] [nQSError: 10058] A general error has occurred. [nQSError: 46036] Internal Assertion: Condition !levelJoinNodes.empty(), file .\Src\SQONPathJoinConstructor.cpp, line 922. (HY000)
Any ideas?
tnks...

Tnks one more time for your response.
I checked everything again and "believe" that everything is ok.
I made a simple report with 3 collumns
One for moth,quantity and -quantity last month-
I checked in the session
-------------------- Logical Request (before navigation):
RqList
MCALENDAR.Μήνας as c1 GB,
Ago(Ποσότητα:[DAggr(MFACT.Ποσότητα by [ MCALENDAR.Μήνας] SB Month)], [Level Month], 1) as c2 GB,
Ποσότητα:[DAggr(MFACT.Ποσότητα by [ MCALENDAR.Μήνας] )] as c3 GB
OrderBy: c1 asc
+++Administrator:2d0000:2d0003:----25/03/2009 20:42:57
-------------------- Query Status: Query Failed: [nQSError: 46036] Internal Assertion: Condition !levelJoinNodes.empty(), file .\Src\SQONPathJoinConstructor.cpp, line 922.
where in greek:
month = Μήνα
quantity = Ποσότητα
year = Έτος
i tried the same report after adding one more column at the end,the column year.
the result was ok.
-------------------- Sending query to database named monthly (id: <<142688>>):
WITH
SAWITH0 AS (select Case when T36043.YEAR_MONTH is not null then DENSE_RANK() OVER ( ORDER BY T36043.YEAR_MONTH ASC NULLS LAST ) end as c1,
T36043.YEAR_MONTH as c2
from
dwadm.MCALENDAR T36043),
SAWITH1 AS (select distinct SAWITH0.c1 + 1 as c1,
SAWITH0.c2 as c2
from
SAWITH0),
SAWITH2 AS (select Case when T36043.YEAR_MONTH is not null then DENSE_RANK() OVER ( ORDER BY T36043.YEAR_MONTH ASC NULLS LAST ) end as c1,
T36043.YEAR_MONTH as c2,
T36043.YEAR_NUM as c3
from
dwadm.MCALENDAR T36043),
SAWITH3 AS (select D1.c1 as c1,
D1.c2 as c2,
D1.c3 as c3
from
(select sum(T34542.MQTY) as c1,
SAWITH2.c2 as c2,
SAWITH2.c3 as c3,
ROW_NUMBER() OVER (PARTITION BY SAWITH2.c2 ORDER BY SAWITH2.c2 ASC) as c4
from
dwadm.MFACT T34542,
SAWITH1,
SAWITH2
where ( T34542.YEAR_MONTH = SAWITH1.c2 and SAWITH2.c1 = SAWITH1.c1 )
group by SAWITH2.c2, SAWITH2.c3
) D1
where ( D1.c4 = 1 ) ),
SAWITH4 AS (select D1.c1 as c1,
D1.c2 as c2,
D1.c3 as c3
from
(select sum(T34542.MQTY) as c1,
T36043.YEAR_MONTH as c2,
T36043.YEAR_NUM as c3,
ROW_NUMBER() OVER (PARTITION BY T36043.YEAR_MONTH ORDER BY T36043.YEAR_MONTH ASC) as c4
from
dwadm.MCALENDAR T36043,
dwadm.MFACT T34542
where ( T34542.YEAR_MONTH = T36043.YEAR_MONTH )
group by T36043.YEAR_MONTH, T36043.YEAR_NUM
) D1
where ( D1.c4 = 1 ) )
select distinct case when SAWITH4.c2 is not null then SAWITH4.c2 when SAWITH3.c2 is not null then SAWITH3.c2 end as c1,
SAWITH3.c1 as c2,
SAWITH4.c1 as c3,
case when SAWITH4.c3 is not null then SAWITH4.c3 when SAWITH3.c3 is not null then SAWITH3.c3 end as c4
from
SAWITH3 full outer join SAWITH4 On nvl(SAWITH3.c2 , 'q') = nvl(SAWITH4.c2 , 'q') and nvl(SAWITH3.c2 , 'z') = nvl(SAWITH4.c2 , 'z')
order by c1, c4
result
month     qty last moth     qty month year
2007/01          661,516.56     2007
2007/02     661,516.56     786,911.55     2007
2007/03     786,911.55     1,282,058.63     2007
2007/04     1,282,058.63     1,389,586.66     2007
2007/05     1,389,586.66     1,981,459.98     2007
2007/06     1,981,459.98     2,393,482.06     2007
2007/07     2,393,482.06     3,181,910.59     2007
2007/08     3,181,910.59     3,736,296.87     2007
2007/09     3,736,296.87     2,208,069.30     2007
2007/10     2,208,069.30     1,536,717.72     2007
2007/11     1,536,717.72     961,400.41     2007
2007/12     961,400.41     985,445.45     2007
2008/01     985,445.45     924,097.26     2008
2008/02     924,097.26     895,622.24     2008
tnksss

Similar Messages

  • AGO function error in Pivot Table

    Hi,
    I have a typical time dimension - year, period, week, date. Most of my facts are available at the most detailed level of date however I have some Budget facts that are only available at the week level. When I combine these facts in a report I am having the below problem.
    I use an ago function at the yearweek level ( I needed this so that I can compare week numbers as oppposed to week date ranges which is how the year ago function works - I believe).
    If I have a total on year in my pivot table I get the below error -
    State: HY000. Code: 10058. [NQODBC] [SQL_STATE: HY000] [nQSError: 10058] A general error has occurred. [nQSError: 22036] Level in AGO function (YearWeek) must be a rollup of (or equivalent to) the measure level (Year). (HY000)
    Has anyone come across this before ?
    Thanks
    Patricia

    I am getting the following error message, not sure how to get rid of this.
    Error Codes: OPR4ONWY:U9IM8TAC:OI2DL65P
    State: HY000. Code: 10058. [NQODBC] [SQL_STATE: HY000] [nQSError: 10058] A general error has occurred. [nQSError: 22036] Level in AGO function (Fiscal Period) must be a rollup of (or equivalent to) the measure level (Fiscal Year). (HY000)
    SQL Issued: SELECT "GL Calendar"."Fiscal Period" saw_0, "Project Manager"."Project Manager Name" saw_1, Project."Project Number" saw_2, Project."Business Area" saw_3, Project."Business Unit" saw_4, Project."Delivery Ownership" saw_5, Project.Programme saw_6, "Fact - Project Cost".Cost saw_7, "Ago Measures"."MAGO Cost" saw_8, "Fact - Project Cost".Cost-"Ago Measures"."MAGO Cost" saw_9, REPORT_SUM(saw_7 BY ), REPORT_SUM(saw_8 BY ), AGGREGATE(saw_9 BY ) FROM "Project - Cost" WHERE ("GL Calendar"."Fiscal Year" = '2011') ORDER BY saw_0, saw_1, saw_2, saw_3, saw_4, saw_5, saw_6

  • AGO function is not returning value in OBIEE 11g

    I have 2 tables FIN_MASTER AND FIN_TRANS. 
    FIN_MASTER  is having the following fields org_code, type_code, vr_no, vr_Date, pay_to .FIN_TRANS is having the columns org_code, type_code, vr_no, acct_code. debit, credit.
    PK of FIN_MASTER is org_code, type_code, vr_no
    PK of FIN_TRANS is org_code, type_code, vr_no, acct_code
    I have imported these two tables in to physical layer and created join between these two tables on the following keys. org_code, type_code, vr_no. Then i dragged these two to BMM layer. Now i created 3 new logical columns in DIM_FIN_MASTER as year, quarter, month. These 3 columns are created using the functions year(vr_date), month(vr_date), quarter_of_year(vr_date). Then i created a dimension heirarchy called TIME( structure type is TIME) from this DIM_FIN_MASTER. and set chronological key on month level. Then I applied the AGO function on debit column with month level. After that, repository is uploaded and BI server is restarted. Now amount is not displaying in AGO function column in analytics. That column is appearing as empty. 
    Please advise me.

    Here is the dirty fix, I'm heading off to sleep.
    If the function CurrentAdId doesn't return a number, then
    this will ensure it returns 0 (prevents your error, may not display
    an ad):
    <CFFUNCTION
    NAME="CurrentAdId"
    ACCESS="Private"
    RETURNTYPE="numeric"
    HINT="For internal use. Returns the Id of the current ad in
    rotation.">
    <!--- Return the adId from the current row of the
    GetAdIds query --->
    <cfset var returnVal = ListGetAt(THIS.AdList,
    THIS.CurrentListPos)>
    <cfif NOT IsNumeric(returnVal)>
    <cfset returnVal = 0>
    </cfif>
    <CFRETURN returnVal>
    </CFFUNCTION>

  • AGO function in OBIEE

    Hi,
    I am creating a new logical column on my fact table and would like to use time series function(AGO) to compare trends a month ago. I have a time dimension too.
    I am using this expression defined in the new logical column properties:
    Ago("HCBF"."Workflow Fact"."Total Case Handlers" , "HCBF"."Time"."Month of Year" , 1)
    I get the following error message:
    [nQSError: 27009] Unresolved identifier: "HCBF"."Time"."Month of Year".
    The function expression appears to be fine, month of year is numeric and total case handlers is a count.
    Can anybody help resolve this error message.
    Thanks

    Hi,
    An issue remains with the AGO function in BI Answers, when I try to run a report in answers with the new column created in the fact table I get the following error in the log file:
    Query Status: Query Failed: [nQSError: 22040] To use AGO function, the storage level of the query ('[Time.Dimension Key]') must be a static level.
    I am selecting month name(Filtered on May 2010) and choosing fact measures(total cases, lastmonthcases). lastmonthcases is the new logical column created using AGO function.
    Does this error look familiar, there is a time dimension key in the logical time dimension table but this is not referenced in the hierarchy.
    Thanks

  • AGO function in OBIEE to display weekly sales data of this year vs last yea

    All,
    I would to create an analysis that display this year sales numbers vs last year by weekly
    I was able to do this at month level by specifying the offset value to 12 to the AGO function in repository.
    I am not able to do at week level.
    Can someone please help?
    Thanks

    Check the W_DAY_D or W_WEEK_D for number of weeks per year, validate the year/4 then its leap year then 53
    Use this in ago function at <<Number of Periods>>.
    I think you have to go for 53 based on these tables some times 54.

  • Level based measure and AGO function

    Hi,
    I've a TIME fact table with a single dimension (TIME_DIM) and a single measure (WORK_DAY).
    WORK_DAY is 1 if the day is a working day, otherwise it is 0.
    This way I can get the number of working days by month, quarter and year.
    (TIME is a logical fact table. WORK_DAY is phsical stored inside the physical TIME_DIM).
    To have WORK_DAY work correctly when used with others fact tables, I've defined it as a level based measure, where all the dimensions are set to the "total" level, apart TIME_DIM.
    So far so good.
    Then I created a new calculated measure WORK_DAY_PREVIOUS_YEAR.
    I used the AGO function this way: AGO (TIME.WORK_DAY, TIME_HIERARCHY.YEAR, 1)
    No complains by OBIEE, but this column always returns 1...
    I've no trouble using level based measures.
    I've no trouble using AGO.
    But if I mix them together they no longer work.
    Do you experience this issue? Any workaround?
    I'm using OBIEE 11.1.1.5
    Regards,
    Corraod

    What is the month level key column functionality in your time dimension?
    it should be in format of yyyymm but not in just mm ..

  • TODATE and AGO function errors in OBIEE

    I have aggregated time dimension. The time calendar comprises of the following structure with a physical table for each level in the Physical Layer -
    Year -> Half Year -> Quarter -> Month -> Week -> Day
    These tables form the calendar. The RPD works fine until I use the TODATE and AGO based fields in the query.
    When I use the AGO and the TODATE time series function to generate the YTD and PY YTD calculations in the Business Modelling Layer. The system gives the following error on Answers -
    Error Codes: OPR4ONWY:U9IM8TAC:OI2DL65P
    State: HY000. Code: 10058. [NQODBC] [SQL_STATE: HY000] [nQSError: 10058] A general error has occurred. [nQSError: 22045] AGO function is used on a time dimension which is logically snowflaked. (HY000)
    SQL Issued: SELECT "- Retail Calendar"."Year Identifier" saw_0, "- Sales"."Gallons Sold" saw_1, "- Sales"."Profit Amount" saw_2, "- Sales"."Sales Quantity" saw_3, "- Sales"."Retail Sales Amount" saw_4, "- Sales"."Licensee Sales Amount" saw_5, "- Sales"."% Retail Sales Amount" saw_6, "- Sales"."% Licensee Sales Amount" saw_7, "- Sales"."Store Average Profit" saw_8, "- Sales"."Sales Amount - Last Year" saw_9 FROM "Reporting Objects" ORDER BY saw_0
    Let me know what I am doing incorrect here?
    I have checked all the joins to see if the time calendar has been incorrectly joined, but I can't find anything. Please help.

    Below are the physical tables involved -
    Year - TIME_YEAR_DM
    Half Year - TIME_HALF_DM
    Quarter - TIME_QTR_DM
    Month - TIME_MTH_DM
    Week - TIME_WK_DM
    Day - TIME_DAY_DM
    Each of these tables has an Identifier key (for example, YR_IDNT, HALF_IDNT, QTR_IDNT, etc.) that uniquely identifies the rows. I am using these IDNT fields as my hierarchical keys.

  • Ago function with two dates

    I use OBIEE and I got some difficulties with the times dimensions and the function AGO because I used two differents dates for my measure.
    My fact table:
    Ordered ||Delivered||Amount
    12/21/2008 ||1/20/2009||30000
    1/1/2009 ||2/1/2009 ||20000
    1/1/2009 ||3/1/2009 ||10000
    2/1/2009 ||2/1/2009 ||50000
    2/1/2009 ||3/1/2009 ||50000
    12/25/2009||1/15/2010||10000
    1/1/2010 ||2/1/2010 ||25000
    1/1/2010 ||3/1/2010 ||20000
    1/1/2010 ||2/1/2010 ||20000
    1/1/2010 ||3/1/2010 ||15000
    I have two Times Dimensions (Ordered Times AND Delivered Times) and one fact table. I already tried with 2 facts table and 1 time table without success too cause they not link the value for delivered and ordered
    I want to compare the delivered situation (Amount Qty) for the same ordered day last year
    I created a prompt with a date (ex.: Ordered date <=1/1/2010 AND Delivered date >=1/1/2010) and I want to compare one year ago with 1/1/2009 (Ordered Date situation…)
    I wish the results below:
    Delivered Month||Amount||Previous Amount Year Ago
    1 || 10000 || 30000
    2 || 45000 || 20000
    3 || 35000 || 10000
    I got some problem cause I had the 100 000 in my previous year (Ago) and I don’t want it because it was ordered after 1/1/2009 (Previous year). I know that the ago function concerns the Sum of Amount by Delivered Date, Year -1 ...
    Does somebody has a solution for this ? Thank you

    Hi,
    The problem is that I only know the offset values at row level, i.e., when I run my answer report some records use a certain offset value (offset1) while others use another offset value (offset2)
    The offset value can be determinate by a logical column.
    Is it possible?
    Thanks.

  • Ago Function

    Hi all
    I upgraded from 10g to 11g and now my Ago function seems to not be working but i have no idea why.
    The data is basically coming from a table like so:
    AcctNbr EffDate Balance
    12345 01/01/2010 $200
    12345 01/02/2010 $205
    12345 01/01/2011 $206
    12345 01/02/2011 $207
    So you can see that I have a balance amount for each day of the year.
    So What I am trying to do is get the Year Ago amount.
    I have this table joined to a basic time dimension table. Year > Quarter > Month > Week > Day.
    In 10g I created a logical column using this and it worked fine:
    AGO("LDW"."FA_ACCTBAL"."NOTEBAL", "LDW"."DimTime"."Year", 1)
    In 11g I get no results. The log file says it my YAGO Measure is being converted to a null. (converted to NULL because its grain is below query's grain.)
    SELECT s_0, s_1, s_2, s_3, s_4, s_5, s_6 FROM (
    SELECT
    0 s_0,
    "LDW"."FA_ACCTBAL"."EFFDATE" s_1,
    CASE WHEN "LDW"."FA_ACCTBAL"."MAJOR" ='CK' THEN 'Chequing'
    WHEN "FA_ACCTBAL"."MAJOR" ='TD' THEN 'Term Deposit'
    WHEN "FA_ACCTBAL"."MAJOR" ='SAV' THEN 'Savings'
    WHEN "FA_ACCTBAL"."MAJOR" ='MTG' THEN 'Mortgage'
    WHEN "FA_ACCTBAL"."MAJOR" ='CNS' THEN 'Consumer'
    WHEN "FA_ACCTBAL"."MAJOR" ='CML' THEN 'Commercial' END s_2,
    "LDW"."FA_ACCTBAL"."% Chg YAGO Notebal" s_3,
    "LDW"."FA_ACCTBAL"."Chg YAGO Notebal" s_4,
    "LDW"."FA_ACCTBAL"."NOTEBAL" s_5,
    "LDW"."FA_ACCTBAL"."YAGO Notebal" s_6
    FROM "LDW"
    WHERE
    ((FA_ACCTBAL.MAJOR IN ('CML', 'CNS', 'MTG')) AND (((FA_ACCTBAL.EFFDATE IN (current_date-1)) OR (FA_ACCTBAL.EFFDATE IN (current_date-366)))))
    ) djm ORDER BY 1, 3 ASC NULLS LAST
    [2011-03-23T21:51:55.000+00:00] [OracleBIServerComponent] [TRACE:7] [USER-23] [] [ecid: 0000Iv^gxcn99hWjLxMMOA1DX^y9000GDt] [tid: 1b6c] [requestid: 5df40012] [sessionid: 5df40000] [username: Administrator] -------------------- General Query Info: [[
    [2011-03-23T21:51:55.000+00:00] [OracleBIServerComponent] [TRACE:7] [USER-52] [] [ecid: 0000Iv^gxcn99hWjLxMMOA1DX^y9000GDt] [tid: 1b6c] [requestid: 5df40012] [sessionid: 5df40000] [username: Administrator] -------------------- Expression 'Ago(NOTEBAL:[DAggr(FA_ACCTBAL.NOTEBAL)], [Level Year], 1)' converted to NULL because its grain is below query's grain.
    [2011-03-23T21:51:55.000+00:00] [OracleBIServerComponent] [TRACE:7] [USER-52] [] [ecid: 0000Iv^gxcn99hWjLxMMOA1DX^y9000GDt] [tid: 1b6c] [requestid: 5df40012] [sessionid: 5df40000] [username: Administrator] -------------------- Expression 'Ago(NOTEBAL:[DAggr(FA_ACCTBAL.NOTEBAL)], [Level Year], 1)' converted to NULL because its grain is below query's grain.
    [2011-03-23T21:51:55.000+00:00] [OracleBIServerComponent] [TRACE:7] [USER-52] [] [ecid: 0000Iv^gxcn99hWjLxMMOA1DX^y9000GDt] [tid: 1b6c] [requestid: 5df40012] [sessionid: 5df40000] [username: Administrator] -------------------- Expression 'Ago(NOTEBAL:[DAggr(FA_ACCTBAL.NOTEBAL)], [Level Year], 1)' converted to NULL because its grain is below query's grain.
    [2011-03-23T21:51:55.000+00:00] [OracleBIServerComponent] [TRACE:7] [USER-52] [] [ecid: 0000Iv^gxcn99hWjLxMMOA1DX^y9000GDt] [tid: 1b6c] [requestid: 5df40012] [sessionid: 5df40000] [username: Administrator] -------------------- Expression 'Ago(NOTEBAL:[DAggr(FA_ACCTBAL.NOTEBAL)], [Level Year], 1)' converted to NULL because its grain is below query's grain.
    [2011-03-23T21:51:55.000+00:00] [OracleBIServerComponent] [TRACE:7] [USER-52] [] [ecid: 0000Iv^gxcn99hWjLxMMOA1DX^y9000GDt] [tid: 1b6c] [requestid: 5df40012] [sessionid: 5df40000] [username: Administrator] -------------------- Expression 'Ago(NOTEBAL:[DAggr(FA_ACCTBAL.NOTEBAL)], [Level Year], 1)' converted to NULL because its grain is below query's grain.
    [2011-03-23T21:51:55.000+00:00] [OracleBIServerComponent] [TRACE:7] [USER-52] [] [ecid: 0000Iv^gxcn99hWjLxMMOA1DX^y9000GDt] [tid: 1b6c] [requestid: 5df40012] [sessionid: 5df40000] [username: Administrator] -------------------- Expression 'Ago(NOTEBAL:[DAggr(FA_ACCTBAL.NOTEBAL)], [Level Year], 1)' converted to NULL because its grain is below query's grain.
    Any idea what has happened here?

    Hi,
    I'm facing the same issue. Did you find a solution for this problem?
    Thx
    Jan

  • Ago function not working properly with time dimension

    Hi all,
    In OBIEE 11g we have the following issue:
    We have a fact which contains transaction's sum for the last day of each month.
    We use ago function like this : ago(<fact.column>,1)When we want to see the result from the time hierarchy the fact with the ago function works properly for quarters and half years but for the months shorter than the previous one can't take the right value /equals to NULL/.
    For example :
    sum for YEAR: 10000
    sum for Half Year 1 : 4000
    sum for Half Year 2 : 6000
    sum for Quarter 1: 1000
    sum for Quarter 2: 3000
    sum for Quarter 3: 2000
    sum for Quarter 4: 4000
    sum for month... - here the ago function is confused with the last day of previous month if it's shorter than itself: For example - for march it cannot take value for end of February ..etc.
    Any ideas?
    Thanks in advance,
    Alexander.

    Hi,
    I faced some issues previously with month, so I used month+year as id instead of month id. Instead of tragging month id 1-12 into your hierarchy, create another column into your table like month+year - 201201, 201202 which creates unique month_year_id and drag this column into month level instead of month_id. So the problem might be solved.
    Regards,
    Anitha.B

  • Todate and AGO function in both Fiscal yer and Calendat year

    I want to create report having column with Todate and ago function formulas. I have to display the facts both in fiscal year and calendar year... how to do this? What value shall i choose in the content tab for the time dimension in this case? My fact table has values for each day...
    I am able to display only columns with fiscal year in time series function with timeDim.fiscal_year... similarly only calendar year with time seriese with calendar year columns... In my hierarchy, I have two branches for Time Dim...one is Fiscal and other is Calendar... and they both are at same level
    Help!

    I have an answer for you...Since you have a correctly defined time dimension (with branches - Fiscal Year->Fiscal Qtr ->Fiscal Month->Fiscal Week and Calendar Year->Cal Qtr->Cal Month->Cal Week->Cal that I assume have the common Child level - Day...) - you should have no problem building time-series logical columns
    In TODATE and AGO you must use relevant Time Dim Level - for example:
    AGO (Column, TimeDim.FiscalYear, 1) = would give you data for previous fiscal year
    AGO (Column, TimeDim.CalYear, 1) = would give you data for previous calendar year
    same goes for qtrs, months, weeks
    If you want to take it to the lowest level Day - it won't matter if you pick Fiscal or Calendar since it's essentially the same.
    I hope this is helpful

  • Ago Function gives NULL values

    Friends,
    I need help regarding Ago function available at column level in BMM.
    I am not at all receiving any error. Unfortunately after Ago function implementation i am getting NULL values for that particular column.
    Here are my implementation steps for getting "Month Ago" values for "Amout column" in Sales table.
    1. i have created Time dimension hieararchy.
    It is like this - Year - > Quarter - > Month - > Actual Date
    2. Defining Keys and making them chronological are also done.
    3. Actual Date is the PK and joins with my Fact.
    4. "Amount" Column in Fact is set to 'Sum'.
    5. i created a new column named " Amount Month Ago" and set following logic to its expression builder.
    Ago("AR Reserve"."Fact - AR Reserve Calc Dtl"."Amount" ,  "AR Reserve"."Dim - TimeDim"."Month" , 1)
    I have not received any error.
    Once i pull this new column (Amont Month Ago) into the presentation layer and try to create a Tabular report. It is not showing any value. Other column values are OK as expected.
    I know i am missing something important here.
    Requesting you all to guide me. I am working on OBIEE 10g.

    Hi,
    Hitting a 'NULL' value in the report hints me of some issues with the Hierarchy/Content setup of the LTS. I suggest you to enable level 7 log and see if the AGO(<measure>) is being nullified by the BI Server due to incompatible facts.
    If it is, then you got to revisit your Period hierarchy and content settings in the Fact LTS. Some generic suggestions are,
    1. Create hierarchies for all dimensions available.
    2. Set the granularity of the sources in the fact to all available hierarchies.
    Hope this helps.
    Thank you,
    Dhar

  • Proper use of AGO function

    Dear All,
    I have to report on cubes that have their data aggregated on the day level. Standard reporting works fine. Now I want to do some historic comparison. I therefore created some calculation columns in Administrator, e.g.
    AGO(fact, dim_time.month,1)
    which shall give me the value of the fact just one month (and quarter and year in other columns) ago. If I report this on a daily basis, e.g day, fact, fact_month_ago I can see the report and calculation is correct, fact_month_ago is exactly the sam as fact from one month ago.
    But now I start to aggregate such a report to the month level (in Answers). So the report is month, fact, fact_month_ago. Now, some months are correct some are different. Looking deeper into this I assume this belongs to the number of days in a month. February has 28 days, so in the monthly report 28 lines are aggregated, even for the fact_month_ago table.
    In this case month_ago is January, which has 31 days, but 29 to 31 are not aggregated, as the report goes only from feb 1st to feb 28th.
    In the end the sum of the month_ago calculation is unequal to the sum of the last month.
    Does this mean I cannot use AGO function to calcualte things like that, or is there any trick or other function that would help here ?
    Regards,
    Knut

    You can have problem if you are going in a lower level such as week or date but if you stay on the month level, normally you must have the good result.
    Check here to have an example of this problem:
    http://obiee101.blogspot.com/2009/01/obiee-leap-year-challenge.html
    Otherwise, check your set up:
    http://gerardnico.com/wiki/dat/obiee/bi_server/design/dimension/time_dimension_parameters_for_ago_and_td_functions#timecalender_dimension_design_verification
    See each punt that are listed under the paragraph Time/Calender Dimension Design Verification. A good time dimension is the key for a good result with the time series function.
    Cheers
    Nico

  • How can I use AGO function with aggregates

    I use aggregated table and a detailed table
    I have time hierarchy whose lowest level is date. For sales I
    created year-to-year (YTD) comparision with AGO function. It
    works OK.
    BUT, I want to speed it up, so I used aggregates. I created new
    time table grouped by MONTH and corresponding agg_SALES table. I
    created connections and mapped the fields. It works OK for
    normal queries, that means my queries at month level use
    aggregates. BUT when I add YTD measures or YAGO measure in query, it goes to
    then SALES table whose grain is date.
    How can I force BI Server to use AGO function with aggregates?
    I use OBIEE 10.1.3.4
    Thank you
    Ishaq
    (Question originally taken from ITtoolbox and posted by Gorazd)
    SORRY - THE QUESTION IS IN THE WRONG FORUM
    Edited by: ishaq12 on Nov 19, 2008 12:21 PM

    Hi,
    you can not do it in Administration tool but you can set agregation for that field in Answers (fx->Agregation rule->Sum) and effect should be the same. Note that it may not work properly on Oracle 10g as generated SQL is not 100% recognized, on 11g everything is OK (at least in my case).
    Regards,
    Marko
    Edited by: user10449532 on 2008.11.19 06:29

  • Using aggregates in AGO function

    Hello!
    I want to use AGO function that would use aggregates.
    Requirements:
    I want to be able to make year-to-year comparision, where I compare period based on specific dates
    This same year-to-year comparision I want to speed up using aggregated fact tables grouped on months level.
    I created aggregated SALES table and aggregated time table (group by month) and it works OK with normal queries. It uses aggregated tables when apropriate (possible)
    When I use AGO function it always goes to fact table, which time dimension is at lowest level (i.e. date). How can I force BI server to use monthly aggregated tables in AGO function?
    Any tip, workaround?
    BI server 10.1.3.2.1
    Thank you,
    Gorazd
    Message was edited by:
    gor
    Message was edited by:
    gor

    I tested it and in case of AGO function it goes to detailed fact table SALES instead on agg_sales. Any tips?
    Gorazd
    Message was edited by:
    gor

Maybe you are looking for