DP time bucket conversion

Hello,
For statistical reason, our statistical forecast is made with monthly time bucket (Gregorian calendar). The business requirement is to have converted the results in fiscal period month. After conversion based on the DP logic we are loosing 5-7 % from our accuracy.
How we can fix that difference?
Any idea would be appreciated.
Thanks, Marius  

Marius,
If you are forecasting at an aggregated level (most companies do), and you have chosen the conversion method of disaggregating months down to days, and then re-aggregating back up to fiscal periods, then 5-7% difference is not unreasonable.  You might be able to achieve better precision by tweaking a few things, or by creating a custom ABAP conversion routine.
Seriously, though, how much effort do you really want to spend on this?  High levels of precision in a forecast are generally inconsequential.  Remember, it is just a forecast, which is a prediction of the future.  When you consider the most common forecast KPI, ( forecast vs actual), a world class forecasting process will only yield 80% accuracy.  In many companies, using a forecast that is 'imprecise' by 5% is generally not worth the time and money to improve.
Best regards,
DB49

Similar Messages

  • Key figure display in planning book with respect to Time bucket profile

    Hi,
    I am loading a key figure to planning area from the info cube for the current month. When I review the key figure in planning book with monthly time bucket profile it shows 85 for the current month. In the same planning book with weekly bucket profile, it shows 55 from the current week and future weeks and the remaining 30 goes into the past weeks of the current month.
    How to make the total quantity 85 to show in the current and future weeks only.
    thanks and regards
    Murugesan

    Hi Murugesan,
    Within the Planning Area, the data is stored at the lowest level granularity that you maintain in storage bucket profile. Then during display, system will decide what data to show depending on what kind of time bucket profile you use in the planning view, and the time based disaggregation that you maintain for Key Figure.
    In this below case, what time characteristic do you have in cube? Is it date, week or month?
    If it's date, check how much KF data is maintained on the dates which belong to week which has days both in this month/last month e.g. if I talk about Dec 2011, how much data is stored 1,2,3 & 4 th of Dec, 2011.
    This data would appear in Dec in monthly view, but in week, it would appear in the week starting 28th November.
    If data is maintained in cube in weeks, then you need to calculate how time based disaggregation would show it to you in months.
    If it's months, then you would need to find out how much data would go to the days in the past week of the month.
    The time based disaggregation may be causing you some issues, but in that case, I would not expect 30 out of 85 to go in the past week, unless you have data in cube in days.
    Data shown in weekly view for week starting 28th Nov should ideally be a small proportion of 85, unless you are using a time stream/fiscal year variant, due to which most of December is in holidays.  The only other exception I can think of is that you have data in teh days mentioned above.
    It would be best to help the business understand this disaggregation logic, rather than thinking of manipulating the data to shift to a later week.
    If this logic doesn't explain your situation, then please provide the date/week/month at which you have data in cube, and what quantity.
    Thanks - Pawan

  • I have the converter to change pdf to word. Works sometime, other time says conversion is taking too long. Can I get what it has translated so I can use the parts?

    I have the converter to change pdf to word. Works sometime, other time says conversion is taking too long. Can I get what it has translated so I can use the parts?

    I would ask in the ExportPDF forum, Adobe ExportPDF (read only) (assuming that's the service to which you have subscribed). This is the Reader one.

  • How to Add attachments at the time of Conversion

    Hi all ,
    Is it possible to get the attachments at the time of conversion .
    we want to attach a URL to the trancation at the time of data conversion .
    Example:
    if we are uploading the MM data and we want to have a URl as an attachment for every material can we add it at the time of conversion?
    thanks
    Naval

    Using GOS ...

  • A simple list with seven time buckets starting from the date report is run.

    Hi All,
    I am new to ABAP.Recently i have planned to write a progam which will help in planning the delivery of scheduled items.For a given sales Org. it will display all the undelivered,delivered items in a simple list and sort them with delivery dates.it will also provide a summary report at material group.For that i have used selection screen(LIKP-VKORG-Obligatory).when i enter VKORG Details i should get all the delivery docs for which delivery is not done or partially done.For this i have taken one more table LIPS
    (LIPS-MATKL,LIPS-MATNR,LIPS-VBELN,LIPS-POSNR,LIPS-WERKS,LIPS-LFIMG,LIPS-MEINS) for tables LIKP and LIPS VBELN is the key field.when i enter VKORG Data in the selection screen ,how can i get the data from LIPS Table.Please explain?
    Also report should be simple list with seven time buckets starting from the date when the report is run.The amount of quantity to be delivered should be displayed under appropriate bucket i.e within the bucket where its delivery date falls in.for ex:If the report is run on Tuesday 15th march 2010 then the start date 1 should be starting date of the week which is Monday 14th March.
    The report will be summarized at Material Group and Material Number.
    Appreciate your help
    Thanks and Regards,
    Shakeer Hussain

    Sorry, sounds too much like a complete requirement you want done for you by the community.
    Please work on it yourself and search for available information before posting specific problems only.
    Thread locked.
    Thomas

  • Time buckets edit/display mode

    Hi
    We have requirement as first two time buckets of the planning book would be in editable mode to the user and from the 3rd bucket onwards the time buckets should be in display mode.
    How can we make the portion of the future time buckets as editable ?
    Thanks & Regards
    KPN Murthy.
    Edited by: p narasimha murthy konidena on Mar 24, 2011 1:59 PM

    KPN
    Look at the SAP help here:
    http://help.sap.com/saphelp_scm70/helpdata/EN/4b/755bee3bf75a18e10000000a421937/frameset.htm
    Look at SAP standard Macros in Macro book 9AEXAMPLES.
    Rishi Menon

  • Time bucket not displaying in planning book

    When I opened planning book, the time bucket is not displaying as it used to be. I tried to select CVC to be displayed in the planning book and got the following error message:
    @5C\QError@     Insufficient information about read period transferred
    @5C\QError@     Error reading data - Planning book cannot be processed further
    Insufficient information about read period transferred
    Message no. /SAPAPO/TSM231
    This internal error should not occur during normal processing. However if this message is displayed, contact your system administration or the SAP hotline quoting the message number.
    Any idea what went wrong?

    Hi Nic,
    Either someone deleted the time bucket profile, or else the periodicities in the time bucket profile are not in line with the periodicity allowed in your planning area.
    Go to Advanced Planning and Optimization-> Demand Planning -> Environment -> Current Settings -> Maintain Time Buckets Profile for Demand Plng and Supply Network Plng, or directly use the tcode "/SAPAPO/TR30".
    Check that your time bucket profiles ZDP_PBP_9Q & ZDP_PBP_5Q exist by using F4 help, and also select them to see what periods are used in the time bucket profile e.g. weeks, days, etc.
    If you don't find the time bucket profile, then they need to be created (based on names, my guess is that *9Q will contain 9 quarters, and *5Q will contain 5 quarters). I expect that issue lies here itself.
    If time buckets profile exist, then check the following:
    1) Go to /sapapo/msdp_admin, select your planning area. See the storage bucket profile name in the "info" tab.
    2) Go to /SAPAPO/TR32  (Periodicities for Planning Area). Give the name of the Storage bucket profile that you see in 1.
    3) Check what periods are used in storage bucket profile.
    4) Storage bucket profile needs to have lower periods maintained than the time bucket profile. e.g. assuming that your time bucket profile has quarters, then you need to have at least quarter maintained in storage bucket profile. Even if you have day or week or month selected in storage bucket profile, it's ok because these are smaller periods than quarters what you have in time bucket profile.
    Also make a note if "post period" is checked. Note the name of "fiscal year variant" if that is maintained along with this tick. I hope you don;t need to go further but if above don;t resolve your issue, then we need to check the fiscal year variant.
    Thanks - Pawan

  • Invalid time buckets profile reference

    Hi :
    I am getting fololowing error while loading data in planning book : " Invalid time buckets profile reference", and unable to laod any data  to planning book
    I have already done consistency check , initailised version , checked time bucket profile , storage bucket profile , and those are fine .There are no time stream id defined either .

    Hi Virendra,
          Check whether your time bucket profile falls same under storage bucket profile. I mean, check your fiscal year variant in storage bucket profile as well as in Time bucket profile. becz you have assigned the storage bucket profile to planning Area, which having fiscal year variant & if you are using different fiscal year variant in Time bucket profile, this mentined error will appear.

  • Unable to save Time Buckets Profile for SNI

    I am getting this weird behavior with the Time Buckets Profile in SNI.
    Whenever I Edit and Save the changes, the Time Bucket Profile reverts back to Years 0000, Months 0000, Weeks 0000 and so on.
    I am using SCM 7.0 with the SP 06.
    I have maintained the necessary Visibility Profiles for the relevant applications.
    It was working until a couple of days and stopped from today. We recently added a custom view under SNI Details.
    Any thoughts?
    Kedar

    Amit
    this is not a Time bucket profile but a storage bucket profile
    you cannot create or maintain it... its standard SAP structure for storing data in technical buckets
    the use of this bucket is in interactive planning book if you want to see how the data is stored(which you seem to have used). This is useful , say, in explaining how the weeks overlapping two months has got the quantity. Again this is a temporary setting and cannot be saved in the planning book
    SCM5.0 makes this a bit accessible by giving this as an option in the drop down icon for "period settings"
    for a related thread see
    Weekly Buckets Incorrectly Split into incomplete weeks!?!?

  • Date Time stamp conversion

    I have a timestamp coming from flat file in the following format YYYY-MM-DD HH:MM:SS.mss(2007-02-21 11:48:31.000)
    to store in Bw I have  created a Object as data type :NUMC and length 14 ,I have conversion routine as times .I was wondering if this will work and also
    can anyone suggest me to covert this as YYYYMMDD and tIme by spliting the field i am getting from flat file.
    Thanks
    Samuel

    Hi Samuel,
    You would need to write a routine for converting.
    Char v_date(8), i_date(18), day(2), month(2), year(2).
    i_date = trans_stru-<your date>.
    day = i_date +  9(2).
    month = i_date + 6(2).
    year = i_date + 5(4).
    concatenate year month day into v_date.
    Result = v_date.
    Bye
    Dinesh
    (Do not forget to assign points!!)

  • Automatic time zone conversion in web dynpro

    We have a WDA running on top of a SAP system that is running on US time. The data from the WDA must be stored in SAP using US time. The WDA will however run in England and needs to display English time to the users. Is there any mechanism or concept in web dynpro that provides automatic time conversion when reading from and displaying time and when writing data back to the database? Some kind of  screen exit or other?
    I am not asking how to convert the value of a single field from one time zone to another, I am hoping for some general solution to the problem of time zones in web dynpro.
    Thanks!

    I don't really have anything useful to say, but couldn't help to add my two cents here...
    Whenever you work in a system that supports users or other parties spread among different time zones you start wishing for some better time zone support. It's not just the user's time zone that is relevant, just imagine for example a service call to a call center in India, where probably the customer's time zone is even more important than the user time zone.
    I know that in principle we have some design issue here, because standalone dates (without any time details) are impossible to convert among different time zones. It would be nice though if the system would provide a framework where any date/time fields can be registered and linked to some time zone field. Changing the time zone field, would automatically update all registered date/time fields.
    Ideally the field for choosing time zones would have the ability to show a list of preferred time zones with (changeable) explanations. One could for example imagine that by default it's populated with the two entries for user time zone and system time zone. Each application should be able to register additional time zones, e.g. sold-to time zone for an order, etc.
    Another (workaround) option I used in applications (for automatically filled read-only fields, like creation date/time) is to settle on one reasonable default time zone (e.g. user) and then have tool tips showing dates/times in other relevant time zones.
    Anyhow, lots of half-baked gibberish, essentially I just wanted to say that it would be nice to have some better time zone support, so I'm voting for that as well (good point Chris). And ideally this would not be limited to web dynpro. I'm sure some smart person should be able to come up with something nice...
    Cheers, harald

  • Grouping by time buckets

    I have an sql that does some counts by hour for the last 24 hours. I wanted to change this to use for the last 4 hours but instead of 24 buckets I wanted 16 buckets of 15 minutes each.
    For the last 4 hours in hour buckets seem to work fine. But for the last 4 hours in 16 buckets, doesn't work fine as it seems to give me the totals for the 1 bucket and repeats the same for the rest of the 15. Could some look let me know what I'm doing wrong here, thanks.
    create table test(dat date, ct1 number, ct2 number);
    declare
    begin
    FOR i IN 1 .. 16
    LOOP
    insert into test values((sysdate - 4/24)+i*1/96, i, i*2);
    commit;
    END LOOP;
    end;
    with dt_minmax as (SELECT trunc(sysdate-4/24,'hh24') AS min_dt,
                   trunc(sysdate,'hh24') AS max_dt
              FROM DUAL),
         dense_days AS
         (SELECT     min_dt + LEVEL / 24 AS dt, min_dt, max_dt, level lvl
                FROM dt_minmax
          CONNECT BY LEVEL <=  4)
    SELECT   dt AS period,
             SUM (s.ct1 + s.ct2) AS BYTES,
             COUNT (TRUNC (s.dat, 'HH24')) AS activity
        FROM dense_days dd LEFT OUTER JOIN test s ON dd.dt = TRUNC (s.dat, 'HH24')
    GROUP BY dd.dt
    ORDER BY dd.dt;
    PERIOD                    BYTES   ACTIVITY
    15-Dec-2007 23:00:00         66          4
    16-Dec-2007 00:00:00        114          4
    16-Dec-2007 01:00:00        162          4
    16-Dec-2007 02:00:00         48          1
    with dt_minmax as (SELECT trunc(sysdate-4/24,'hh24') AS min_dt,
                   trunc(sysdate,'hh24') AS max_dt
              FROM DUAL),
         dense_days AS
         (SELECT     min_dt + LEVEL / 96 AS dt, min_dt, max_dt, level lvl
                FROM dt_minmax
          CONNECT BY LEVEL <=  16)
    SELECT   dt AS period,
             SUM (s.ct1 + s.ct2) AS BYTES,
             COUNT (TRUNC (s.dat, 'HH24')+lvl/96) AS activity
        FROM dense_days dd LEFT OUTER JOIN test s ON dd.dt = (TRUNC (s.dat, 'HH24')+lvl/96)
    GROUP BY dd.dt
    ORDER BY dd.dt;
    PERIOD                    BYTES   ACTIVITY
    15-Dec-2007 22:15:00         18          3
    15-Dec-2007 22:30:00         18          3
    15-Dec-2007 22:45:00         18          3
    15-Dec-2007 23:00:00         18          3
    15-Dec-2007 23:15:00         18          3
    15-Dec-2007 23:30:00         18          3
    15-Dec-2007 23:45:00         18          3
    16-Dec-2007 00:00:00         18          3
    16-Dec-2007 00:15:00         18          3
    16-Dec-2007 00:30:00         18          3
    16-Dec-2007 00:45:00         18          3
    16-Dec-2007 01:00:00         18          3
    16-Dec-2007 01:15:00         18          3
    16-Dec-2007 01:30:00         18          3
    16-Dec-2007 01:45:00         18          3
    16-Dec-2007 02:00:00         18          3
    16 rows selected.

    I'm not sure if ntile would work in this case, since ntile puts them into equal number of buckets. But we want them to be divided by time range as in 15 minutes of 16 buckets fro 4 hours. But if there are more number of records in say one 15 minute cycle, some of these records will be pushed to the next cycle as it tries tp put them in equal buckets.
    SQL> SELECT   bucket, dat
        FROM (SELECT NTILE (16) OVER (ORDER BY dat) bucket, dat
                FROM TEST
               WHERE dat >= trunc(sysdate-4/24,'hh24') and dat <= trunc(sysdate,'hh24')) s
    GROUP BY bucket,dat
    order by bucket;  2    3    4    5    6
        BUCKET DAT
             1 25-Dec-2007 10:32:02
             2 25-Dec-2007 10:32:03
             2 25-Dec-2007 10:32:04
             3 25-Dec-2007 10:32:05
             3 25-Dec-2007 10:38:48
             4 25-Dec-2007 10:47:02
             4 25-Dec-2007 10:53:48
             5 25-Dec-2007 11:02:02
             5 25-Dec-2007 11:08:48
             6 25-Dec-2007 11:17:02
             6 25-Dec-2007 11:23:48
        BUCKET DAT
             7 25-Dec-2007 11:32:02
             7 25-Dec-2007 11:38:48
             8 25-Dec-2007 11:47:02
             8 25-Dec-2007 11:53:48
             9 25-Dec-2007 12:02:02
             9 25-Dec-2007 12:08:48
            10 25-Dec-2007 12:17:02
            10 25-Dec-2007 12:23:48
            11 25-Dec-2007 12:32:02
            11 25-Dec-2007 12:38:48
            12 25-Dec-2007 12:47:02
        BUCKET DAT
            12 25-Dec-2007 12:53:48
            13 25-Dec-2007 13:02:02
            13 25-Dec-2007 13:08:48
            14 25-Dec-2007 13:17:02
            14 25-Dec-2007 13:23:48
            15 25-Dec-2007 13:32:02
            15 25-Dec-2007 13:38:48
            16 25-Dec-2007 13:47:02
            16 25-Dec-2007 13:53:48
    31 rows selected.

  • Time Buckets with multiple Key Figures in Each

    Hello Experts,
    I have seen documents creation of ageing buckets and calculations but not seen any document with ageing bucket having multiple Key Figures part of it.
    User's have used Excel to create Bucket header and using V and H lookups they are populating data in to those columns. Is there a way we can achieve this in Bex Designer? Below is what needed on report which would give user's overview of Current Week, Previous Week, Last 4 weeks and Last 13 weeks. I have attached Snapshot for better Understanding
    From Excel--->
    Target
    Current Week
    Last 4 Weeks
    Utilization
    Billable
    Non-billable
    Non Work
    Utilization
    Billable
    Non-billable
    Non Work
    Utilization
    Billable
    Non-billable
    Non Work
    Thanks
    Sam

    Hi Sam
    If week is your time period analysis then create a structure in columns, above your key figures, based on week selections variable.
    pic example
    I don't think that you will find such a document.
    Thank you
    Yiannis

  • Time Buckets in DP

    Dear All,
    I have the following scenario in APO DP:
    I have two plants: Plant A and Plant B.
    For Plant A (Weekly off is wednesday):
    So Weekly buckets for Oct 08 should look like this
    W1 --> 2nd to 7th Oct
    W2 --> 9th to 14th Oct
    W3 --> 16th to 21st Oct
    W4 --> 23rd to 29th OCt
    And for Plant B (Weekly off is sunday)
    So weekly buckets should look like this
    W1 --> 1st to 4th Oct
    W2 --> 6th to 11th oct...
    continues like this..
    How do i address this in ONE PLANNING AREA?
    Orelse should i create two fiscal year variants and attach to TWO DIFFERENT planning area?

    Hi.
    If I understand your requirement correctly, you are going to need two storage bucket profiles. This being the case, you will need two planning areas.
    Not only will you require two FYVs but also two Time streams to assign to the storage bucket profiles. In the timestreams you can define the order of your workdays in a week.
    Hope this helps.
    M

  • Function module for Time zone conversion.

    Hi SDN,
    Can any one tell is there any FM for conversion of time from GMT to CET or any other time zone
    Regards,
    Rahul

    Try these function modules
    STU3_ALERTS_CONV_LOCAL_2_UTC   Converts local time to a given timezone
    STU3_ALERTS_CONV_UTC_2_LOCAL   Converts local time to a given timezone
    TZ_LOCATION_TIMEZONE                     Time zones: Determine time zone for a location                                                                          (land, region)

Maybe you are looking for