cubic regression with no quadratic term
 

how do i do cubic regression without the quadratic term? how do i tell excel
there is no x^2? i just want it to be ax^3+bx+c

Cyanlce wrote:
how do i do cubic regression without the quadratic term? how do i tell excel
there is no x^2? i just want it to be ax^3+bx+c


I presume you have a column of X and a column of Y data. Add a third
column which is your scoring function that you want to minimize.
Conventionally that would be something like { } = (Y - YourF(x))^2 At
the bottom of that column you sum up all your error function values.

Then you just use Solver to optimize the a, b, c values that you've used
in your scoring function equation to minimize the sum of your error
function values.

Good luck...

Bill

Assuming that yData and xData are columnar data with no empty or
non-numeric cells, then select 3 contiguous cells in a row and array
enter (Ctrl-Shift-Enter)

=LINEST(yData,xData^{1,3})

As documented by Help, the reculting coeffients will be

(cubic, linear, intercept)

which is the opposite order to what you might expect.

Jerry

Cyanlce wrote:

how do i do cubic regression without the quadratic term? how do i tell excel
there is no x^2? i just want it to be ax^3+bx+c

thanks

"Cyanlce" wrote:

how do i do cubic regression without the quadratic term? how do i tell excel
there is no x^2? i just want it to be ax^3+bx+c