DMX

Question

I have this query:

And I would like to use the predict function to introduce lag on my data to evaluate what sales may do next year.

What is the best way to accomplish this?

SELECT
  [LinearR].[Units Sold],
  t.[ProductControlID]
From
  [LinearR]
PREDICTION JOIN
  OPENQUERY([Intellitrak],
    'SELECT
      [ProductControlID],
      [UnitsSold],
      [Yr1975],
      [Yr1976],
      [Yr1977],
      [Yr1978],
      [Yr1979],
      [Yr1980],
      [Yr1981],
      [Yr1982],
      [Yr1983],
      [Yr1984],
      [Yr1985],
      [Yr1986],
      [Yr1987],
      [Yr1988],
      [Yr1989],
      [Yr1990],
      [Yr1991],
      [Yr1992],
      [Yr1993],
      [Yr1994],
      [Yr1995],
      [Yr1996],
      [Yr1997],
      [Yr1998],
      [Yr1999],
      [Yr2000],
      [Yr2001],
      [Yr2002],
      [Yr2003],
      [Yr2004],
      [Yr2005],
      [Yr2006],
      [Yr2007],
      [Yr2008],
      [Yr2009],
      [Yr2010],
      [Yr2011]
    FROM
      [dbo].[vwForModel]
    ') AS t
ON
  [LinearR].[Units Sold] = t.[UnitsSold] AND
  [LinearR].[Yr1975] = t.[Yr1975] AND
  [LinearR].[Yr1976] = t.[Yr1976] AND
  [LinearR].[Yr1977] = t.[Yr1977] AND
  [LinearR].[Yr1978] = t.[Yr1978] AND
  [LinearR].[Yr1979] = t.[Yr1979] AND
  [LinearR].[Yr1980] = t.[Yr1980] AND
  [LinearR].[Yr1981] = t.[Yr1981] AND
  [LinearR].[Yr1982] = t.[Yr1982] AND
  [LinearR].[Yr1983] = t.[Yr1983] AND
  [LinearR].[Yr1984] = t.[Yr1984] AND
  [LinearR].[Yr1985] = t.[Yr1985] AND
  [LinearR].[Yr1986] = t.[Yr1986] AND
  [LinearR].[Yr1987] = t.[Yr1987] AND
  [LinearR].[Yr1988] = t.[Yr1988] AND
  [LinearR].[Yr1989] = t.[Yr1989] AND
  [LinearR].[Yr1990] = t.[Yr1990] AND
  [LinearR].[Yr1991] = t.[Yr1991] AND
  [LinearR].[Yr1992] = t.[Yr1992] AND
  [LinearR].[Yr1993] = t.[Yr1993] AND
  [LinearR].[Yr1994] = t.[Yr1994] AND
  [LinearR].[Yr1995] = t.[Yr1995] AND
  [LinearR].[Yr1996] = t.[Yr1996] AND
  [LinearR].[Yr1997] = t.[Yr1997] AND
  [LinearR].[Yr1998] = t.[Yr1998] AND
  [LinearR].[Yr1999] = t.[Yr1999] AND
  [LinearR].[Yr2000] = t.[Yr2000] AND
  [LinearR].[Yr2001] = t.[Yr2001] AND
  [LinearR].[Yr2002] = t.[Yr2002] AND
  [LinearR].[Yr2003] = t.[Yr2003] AND
  [LinearR].[Yr2004] = t.[Yr2004] AND
  [LinearR].[Yr2005] = t.[Yr2005] AND
  [LinearR].[Yr2006] = t.[Yr2006] AND
  [LinearR].[Yr2007] = t.[Yr2007] AND
  [LinearR].[Yr2008] = t.[Yr2008] AND
  [LinearR].[Yr2009] = t.[Yr2009] AND
  [LinearR].[Yr2010] = t.[Yr2010] AND
  [LinearR].[Yr2011] = t.[Yr2011]

---

foxjazz

Answer 1

I have this query:

And I would like to use the predict function to introduce lag on my data to evaluate what sales may do next year.

What is the best way to accomplish this?

Hi foxjazz,

Did the model use Time Series algorithm? If so, you could call PredictTimeSeries dmx function to predict the future sales data.

http://msdn.microsoft.com/en-us/library/ms174923.aspx is to introduce data mining Time Series algorithm.

http://msdn.microsoft.com/en-us/library/ms132167.aspx is to introduce how to use PredictTimeSeries function with the Time Series algorithm model.

If i misunderstand you, please elaborate the requirement in more details.

Regards,

Jerry

TechNet Subscriber Supportin forum
If you have any feedback on our support, please contact  [e-mail]

Answer 2

Where's my brain: I thought I mentioned I was using Linear Regression.

Musta had to redo this for some reason. Anyway ideas, and how do you get the statistics out?

---

foxjazz