Skip to main content
Topics#Forecast

#Forecast2 discussing

Ask a question...
Joost Breggeman (Senior Director Sales Operations Europe at Chargepoint) asked in #Forecasting
Sep 22, 6:36 AM
Salesforce forecasting submit button missing on product level

We have multiple forecasting types setup. One for Opportunity level forecasting which has a submit button.  

 

We also setup a forecasting type on Opportunity Line Item level. However, we don't have a submit button on product level forecast. Is there a way to fix this?  

 

#Forecasting  #Forecast Type  #Forecast Items  #Forecast

1 answer
  1. Rahul Chauhan (Kcloud)
    Sep 22, 7:06 AM

     The missing Submit button on the Opportunity Product forecast may be expected behavior.

    Salesforce currently states that Forecast Submissions are available only for active forecast types that don't include Product Families or other forecast groups.

    So I would check the configuration of the Opportunity Product forecast type:

    Setup --> Forecasts Settings --> Forecast Types

    Check whether the product-level forecast is using:

    • Object: Opportunity Product
    • Group By: Product Family or another forecast group

    If the forecast type includes Product Family/another grouping, the Submit Forecast option won't be available for that forecast type.

    Also make sure Forecast Submissions is enabled under Setup --> Forecasts Settings. Salesforce says the submission icon appears on the Forecasts page when a forecast type supports submissions.

    So this is likely a forecast-type limitation/configuration, rather than a missing permission or page-layout issue.

Write an answer...
0/9000
Tableau Community Forums Member (Inactive) (Member of the Tableau Community Forums at Salesforce)
Feb 9, 2016, 6:46 AM
24 hour predictive forecast using r with tableau

I've been trying to come up with a 24 hour predictive forecast of sales in tableau with r based on a past years sales.  What I want is to be able to bring in a new 3 hours of data and try to predict how the next 21 hours will go.  I've tried numerous approaches like the ones below, with no luck.  I think the main problem I'm having is forecasting past the end of the date field in tableau.  I found an example using quarters at this link https://boraberan.wordpress.com/2014/01/19/using-r-forecasting-packages-from-tableau/, but I'm having no luck duplicating it with hours.

 

Forecasting within datetimes in my datasource isn't a problem, but forecasting past the last datetime say '2015-09-01 11:00:00' is giving me a headache.  Also dealing with it at the hour level seems to be trickier than say 12 months.  Any tips greatly appreciated.

 

Attempt1

 

    SCRIPT_REAL("library(forecast);

    library(forecast);

    trainingdata<-ts(.arg1,h=24)

    fit <- ets(trainingdata)

    testdata<-ts(.arg3,h=24)

    fit2 <- ets(testdata, model=fit)

    fcast <- forecast(fit, h=24)

    onestep <- fitted(fcast)",

    [SaleCount],[PeriodsToForecast],[RecentSales])

 

Attempt2

 

    SCRIPT_REAL("library(forecast);

  #⌗Train Model

  ts1Train <- ts(.arg1, frequency=24)

  ets1 <- ets(ts1Train,model='MMM')

  ##Forecast Model

  fcast <- forecast(ets1, h=.arg2[1]);

  n<-length(.arg1);

  append(.arg1[(.arg2[1]+1):n],fitted.values(fcast), after = n-.arg2[1])",

  IFNULL([SaleCount],0),[PeriodsToForecast])

2 answers
  1. Tableau Community Forums Member (Inactive) (Salesforce)
    Feb 9, 2016, 4:04 PM

    Hi Bora, thank you for getting back to me so quickly. 

    My data is pretty simple.  I have counts of sales with a datetime field when they occurred.  So basically SalesCount 3 , DateTime '2015-09-01 14:00:00'. 

    Then in tableau my column is set to hour(DateTime) and my row is sum(SalesCount).

    I train my forecast on the entire years worth of data, but I'm only looking to forecast the next 21 hours.

    The end goal is to show the first 3 hours of real data with the next 21 hours forecasted.  Also it would be a different 3 hours throughout the day depending on when the datasource is refreshed.

     

          From: Bora Beran <tableaucommunity@tableau.com>

     

    Sent: Tuesday, February 9, 2016 12:09 AM

    Subject: Re:  - 24 hour predictive forecast using r with tableau

     

    ⌗yiv2377922331 * ⌗yiv2377922331 a

    ⌗yiv2377922331 body {font-family:Helvetica, Arial, sans-serif;}⌗yiv2377922331 ⌗yiv2377922331 h1, ⌗yiv2377922331 h2, ⌗yiv2377922331 h3, ⌗yiv2377922331 h4, ⌗yiv2377922331 h5, ⌗yiv2377922331 h6, ⌗yiv2377922331 p, ⌗yiv2377922331 hr {}⌗yiv2377922331 .yiv2377922331button td {}
    |
     
      Tableau Support Community 
     
    |
    24 hour predictive forecast using r with tableau
    reply from Bora Beran in Forums - View the full discussionWhat does the error message look like? Also can you share your data source even if it has dummy data? The trick is extending the axis to accommodate for the new data, so data structure is crucial to avoid errors. One thing I noticed reading the script is that you are forecasting 24 periods while in the description you said you wanted 21. Also with 3 hours of training data you won't be able to get a forecast that has a 24 hour cycle. You will need more data but you can always hide the data you don't want to display using a table calc filter.
    Reply to this message by replying to this email, or go to the message on Tableau Support Community
    Start a new discussion in Forums by email or at Tableau Support Community
    Following 24 hour predictive forecast using r with tableau in these streams: Inbox
     
      This email was sent by Tableau Support Community because you are a registered user.
    You may unsubscribe instantly from Tableau Support Community, or adjust email frequency in your email preferences  |
     
      |
Write an answer...
0/9000