X
PrevPrev Go to previous topic
NextNext Go to next topic
Last Post 29 Jul 2015 12:21 PM by  anon
Help with Regress.pro or Regress2.pro
 1 Replies
Sort:
You are not authorized to post a reply.
Author Messages

anon



New Member


Posts:
New Member


--
29 Jul 2015 12:21 PM
    So i wanted to use Regress.pro. Usually when we fit a data, we have constant and coefficients. Regress has both. However, there is only error for Coefficients but not for the constant. For example, if we have Y = mX + C, we have errors for both m and C, or Y = m1 X1 + m2 X2 + C and the error for m1, m2 and C. Regress doesnt have error for C. I dont seem to understand why we do not have one. Then i moved to Regress2 hoping for a better fit. However, Regress2 doesnt have constant factor and has been fitting along the lines of Y = mX or Y = m1 X1 + m2 X2. Am i missing something or not understanding it.

    Deleted User



    New Member


    Posts:
    New Member


    --
    30 Jul 2015 09:33 AM
    Well, i couldnt find any answer to why the constant term doesnt have an error in regress.pro. However, we can get around it in Regress2.pro . For Y = aX1 + bX2 in regress2, we just add in Y = aX1 + bX2 + cX0 where X0 is just a vector of the same length as X1 and consists of all 1s. This way C is the constant term and we get the error for C. Cheers -Sam
    You are not authorized to post a reply.