I have a dataset with 3 variables (X,Y and Z) and I want to find the best estimates for the constants a,b,c & d. I have been looking into multiregression analysis, but that does not seem to work for this more complicated formula.

Y = (a + b * Z) * X^(c + d * Z)

Someone got any leads?


You are trying to linearize the problem. It's not possible. You have to use non-linear regression in this case.

  • $\begingroup$ How would I approach this. I do know how to perform non-linear regression, however I do not have experience with 3 variables and would not know where to start. $\endgroup$ – Jens de Bruijn Sep 11 '14 at 13:51
  • $\begingroup$ What stat package are you using? $\endgroup$ – Aksakal Sep 11 '14 at 13:52
  • $\begingroup$ Python, but I could use any $\endgroup$ – Jens de Bruijn Sep 11 '14 at 13:58
  • $\begingroup$ Take a look at this R doc. It has an example in it. and a theory if you want to implement the thing in your code manually. $\endgroup$ – Aksakal Sep 11 '14 at 14:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.