Go Back   PC Gamer Forums > Off topic

    Reply
     
    Thread Tools Display Modes
      #1   Report Post  
    Old 11-20-2012, 01:56 PM
    Spatula Spatula is offline
    Senior Member
     
    Join Date: Jun 2010
    Posts: 2,916
    Default Statistics, Data residuals and DoE software

    Anyone good with statistics? I'm struggling to get my head around data transformation wrt data residuals.

    Specifically in the context of computational modelling using a DOE software package, but any help on data transformation and residuals would be really useful....

    ...here's hoping! :-)
    __________________
    ----------------------------------------------------------------------------------


    -------
    An irreverent look at gaming, and film; www.TheFunTurret.com
    Reply With Quote
      #2   Report Post  
    Old 11-20-2012, 02:21 PM
    Burglar Burglar is offline
    Senior Member
     
    Join Date: Jun 2010
    Location: Essex Massive
    Posts: 1,454
    Default

    The one thing you've got to remember about data residuals and DoE software is...OH LOOK OVER THERE IT'S GABE NEWELL STANDING RIGHT BEHIND YOU!

    *flees*
    Reply With Quote
      #3   Report Post  
    Old 11-20-2012, 02:46 PM
    Spatula Spatula is offline
    Senior Member
     
    Join Date: Jun 2010
    Posts: 2,916
    Default

    :-)


    /tenchar
    __________________
    ----------------------------------------------------------------------------------


    -------
    An irreverent look at gaming, and film; www.TheFunTurret.com
    Reply With Quote
      #4   Report Post  
    Old 11-20-2012, 10:48 PM
    JackRabbit JackRabbit is offline
    Senior Member
     
    Join Date: Jun 2010
    Posts: 3,601
    Default

    Sorry, my statistical knowledge pretty much ends with a t-test.
    Reply With Quote
      #5   Report Post  
    Old 11-21-2012, 08:05 AM
    Spatula Spatula is offline
    Senior Member
     
    Join Date: Jun 2010
    Posts: 2,916
    Default

    Yeah, the same with most of my coeagues.
    __________________
    ----------------------------------------------------------------------------------


    -------
    An irreverent look at gaming, and film; www.TheFunTurret.com
    Reply With Quote
      #6   Report Post  
    Old 11-21-2012, 08:14 AM
    alberthammond alberthammond is offline
    Senior Member
     
    Join Date: Feb 2011
    Location: Greater Mancunia/Scouseland
    Posts: 1,001
    Default

    Lies! Damn Lies! All of it!
    __________________
    Om - Thats a funny thing, winners never talk of Glorious Victories. That's because they're the ones who see what the battlefield looks like afterward. It's only the losers who have Glorious Victories.

    T Pratchett, Small Gods.
    Reply With Quote
      #7   Report Post  
    Old 11-21-2012, 11:52 AM
    Daniel Daniel is offline
    Senior Member
     
    Join Date: Aug 2010
    Location: People's Repuplic of Treacle Mine Road
    Posts: 3,972
    Default

    I am not entirly sure what you mean by data residuals

    Do you want to know how far each data point on a graph lies from the theoretical true value?

    I could be completly barking up the wrong tree
    __________________
    “There has seldom if ever been a shortage of eager young males prepared to kill and die to preserve the security, comfort and prejudices of their elders, and what you call heroism is just an expression of this simple fact; there is never a scarcity of idiots.” - Ian Banks
    Reply With Quote
      #8   Report Post  
    Old 11-21-2012, 12:03 PM
    Spatula Spatula is offline
    Senior Member
     
    Join Date: Jun 2010
    Posts: 2,916
    Default

    i'm wanting to know how to know when data needs transforming. When to recognise the residual spread that indicates transforming is required and then which transformation to try (short of trial and error).

    The software i'm using can throw up hundreds of models that need reviewing, so i need to really understand this so i can quickly, and accurately scan through the model outputs to check they're accurate before i use them to make any extrapolations.

    clear as mud :-)
    __________________
    ----------------------------------------------------------------------------------


    -------
    An irreverent look at gaming, and film; www.TheFunTurret.com
    Reply With Quote
      #9   Report Post  
    Old 11-21-2012, 03:48 PM
    Daniel Daniel is offline
    Senior Member
     
    Join Date: Aug 2010
    Location: People's Repuplic of Treacle Mine Road
    Posts: 3,972
    Default

    Is that where you do something to the equation you are plotting to make it a straight line linear graph. If you had an exponential curve if you take a log of the independent variable you get a straight line. Or if an independent variable is a quadratic you can square root it.

    That is all I know and is probably not of much help so... good luck
    __________________
    “There has seldom if ever been a shortage of eager young males prepared to kill and die to preserve the security, comfort and prejudices of their elders, and what you call heroism is just an expression of this simple fact; there is never a scarcity of idiots.” - Ian Banks
    Reply With Quote
      #10   Report Post  
    Old 11-21-2012, 06:42 PM
    JackRabbit JackRabbit is offline
    Senior Member
     
    Join Date: Jun 2010
    Posts: 3,601
    Default

    Have you used a......
    Pie chart?
    ¬¬
    Reply With Quote
    Reply

    Thread Tools
    Display Modes

    Posting Rules
    You may not post new threads
    You may not post replies
    You may not post attachments
    You may not edit your posts

    BB code is On
    Smilies are On
    [IMG] code is On
    HTML code is Off

    Forum Jump