+ Reply to Thread
Results 1 to 5 of 5

how to use multiple regression

  1. #1
    Dave in naigeria
    Guest

    how to use multiple regression

    Please can anyone help me solve this problem.

    Use linest function to estimate what A,B,C & D will be, Compare them.

    Y= a+bx+cx^2+dx^3

    assum x=0, x=1, .................. x= 100

    Calculate the corresponding value of Y. Choose any coefficient for the
    constant.

    Thanks and regards,
    Dave

  2. #2
    Conrad Carlberg
    Guest

    Re: how to use multiple regression

    As written, your post is a little puzzling.

    To use LINEST() you're going to need to start with both a series of X values
    and a series of Y values. So you won't "Calculate the corresponding value of
    Y," but you can use the regression coefficients and the constant to get
    estimates of Y. If what you're really after is the Y estimates, the TREND()
    function is more straightforward than LINEST() because it calculates the
    equation and applies it to your X values to return the Y estimates. But if
    you use LINEST(), and assuming you have Y in column A, X in column B, X^2 in
    column C and X^3 in column D, start by selecting a blank range four columns
    wide and five rows high, type this formula:

    =LINEST(A1:A100,B1:D100,,TRUE)

    and array-enter it with CTRL-SHIFT-ENTER instead of just ENTER. You can also
    square and cube X in the arguments:

    =LINEST(A1:A100,B1:B100^{1,2,3},,TRUE)

    What do you want to compare A,B,C & D to? Zero? Something else? Use the
    standard errors.

    You can force the constant to zero using LINEST's third argument but in
    general it's a bad idea.

    --
    C^2
    Conrad Carlberg

    Excel Sales Forecasting for Dummies, Wiley, 2005


    "Dave in naigeria" <Dave in [email protected]> wrote in
    message news:[email protected]...
    > Please can anyone help me solve this problem.
    >
    > Use linest function to estimate what A,B,C & D will be, Compare them.
    >
    > Y= a+bx+cx^2+dx^3
    >
    > assum x=0, x=1, .................. x= 100
    >
    > Calculate the corresponding value of Y. Choose any coefficient for the
    > constant.
    >
    > Thanks and regards,
    > Dave




  3. #3
    Dave in naigeria
    Guest

    Re: how to use multiple regression

    Thanks Conrad.
    The equation is actually not written well x^2 means x squared and x^3 means
    x cube.so the equation is

    Y= a + bx + cxsquared + dxcube

    "Conrad Carlberg" wrote:

    > As written, your post is a little puzzling.
    >
    > To use LINEST() you're going to need to start with both a series of X values
    > and a series of Y values. So you won't "Calculate the corresponding value of
    > Y," but you can use the regression coefficients and the constant to get
    > estimates of Y. If what you're really after is the Y estimates, the TREND()
    > function is more straightforward than LINEST() because it calculates the
    > equation and applies it to your X values to return the Y estimates. But if
    > you use LINEST(), and assuming you have Y in column A, X in column B, X^2 in
    > column C and X^3 in column D, start by selecting a blank range four columns
    > wide and five rows high, type this formula:
    >
    > =LINEST(A1:A100,B1:D100,,TRUE)
    >
    > and array-enter it with CTRL-SHIFT-ENTER instead of just ENTER. You can also
    > square and cube X in the arguments:
    >
    > =LINEST(A1:A100,B1:B100^{1,2,3},,TRUE)
    >
    > What do you want to compare A,B,C & D to? Zero? Something else? Use the
    > standard errors.
    >
    > You can force the constant to zero using LINEST's third argument but in
    > general it's a bad idea.
    >
    > --
    > C^2
    > Conrad Carlberg
    >
    > Excel Sales Forecasting for Dummies, Wiley, 2005
    >
    >
    > "Dave in naigeria" <Dave in [email protected]> wrote in
    > message news:[email protected]...
    > > Please can anyone help me solve this problem.
    > >
    > > Use linest function to estimate what A,B,C & D will be, Compare them.
    > >
    > > Y= a+bx+cx^2+dx^3
    > >
    > > assum x=0, x=1, .................. x= 100
    > >
    > > Calculate the corresponding value of Y. Choose any coefficient for the
    > > constant.
    > >
    > > Thanks and regards,
    > > Dave

    >
    >
    >


  4. #4
    Conrad Carlberg
    Guest

    Re: how to use multiple regression

    >The equation is actually not written well x^2 means x squared and x^3 means
    > x cube.so the equation is
    >
    > Y= a + bx + cxsquared + dxcube


    That much was clear.
    --
    C^2
    Conrad Carlberg

    Excel Sales Forecasting for Dummies, Wiley, 2005


    "Dave in naigeria" <[email protected]> wrote in
    message news:[email protected]...
    > Thanks Conrad.
    > The equation is actually not written well x^2 means x squared and x^3

    means
    > x cube.so the equation is
    >
    > Y= a + bx + cxsquared + dxcube
    >
    > "Conrad Carlberg" wrote:
    >
    > > As written, your post is a little puzzling.
    > >
    > > To use LINEST() you're going to need to start with both a series of X

    values
    > > and a series of Y values. So you won't "Calculate the corresponding

    value of
    > > Y," but you can use the regression coefficients and the constant to get
    > > estimates of Y. If what you're really after is the Y estimates, the

    TREND()
    > > function is more straightforward than LINEST() because it calculates the
    > > equation and applies it to your X values to return the Y estimates. But

    if
    > > you use LINEST(), and assuming you have Y in column A, X in column B,

    X^2 in
    > > column C and X^3 in column D, start by selecting a blank range four

    columns
    > > wide and five rows high, type this formula:
    > >
    > > =LINEST(A1:A100,B1:D100,,TRUE)
    > >
    > > and array-enter it with CTRL-SHIFT-ENTER instead of just ENTER. You can

    also
    > > square and cube X in the arguments:
    > >
    > > =LINEST(A1:A100,B1:B100^{1,2,3},,TRUE)
    > >
    > > What do you want to compare A,B,C & D to? Zero? Something else? Use the
    > > standard errors.
    > >
    > > You can force the constant to zero using LINEST's third argument but in
    > > general it's a bad idea.
    > >
    > > --
    > > C^2
    > > Conrad Carlberg
    > >
    > > Excel Sales Forecasting for Dummies, Wiley, 2005
    > >
    > >
    > > "Dave in naigeria" <Dave in [email protected]> wrote in
    > > message news:[email protected]...
    > > > Please can anyone help me solve this problem.
    > > >
    > > > Use linest function to estimate what A,B,C & D will be, Compare them.
    > > >
    > > > Y= a+bx+cx^2+dx^3
    > > >
    > > > assum x=0, x=1, .................. x= 100
    > > >
    > > > Calculate the corresponding value of Y. Choose any coefficient for the
    > > > constant.
    > > >
    > > > Thanks and regards,
    > > > Dave

    > >
    > >
    > >




  5. #5
    Jerry W. Lewis
    Guest

    Re: how to use multiple regression

    Conrad already explained how to fit a cubic polynomial with LINEST. An
    alternative to using the LINEST coefficients to predict y values at 1 to
    100 would be

    =TREND(known_ys,known_xs^{1,2,3},ROW(A1:A100))

    array entered.

    Jerry

    Dave in naigeria wrote:

    > Please can anyone help me solve this problem.
    >
    > Use linest function to estimate what A,B,C & D will be, Compare them.
    >
    > Y= a+bx+cx^2+dx^3
    >
    > assum x=0, x=1, .................. x= 100
    >
    > Calculate the corresponding value of Y. Choose any coefficient for the
    > constant.
    >
    > Thanks and regards,
    > Dave



+ Reply to Thread

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts

Search Engine Friendly URLs by vBSEO 3.6.0 RC 1