I am poor in statistics. I did a multiple linear regression in python by using statsmodels. I know that I can transform a non-linear relation into a linear relation in different ways using this model. Combinations of different independent variables are shown in below:

smf.ols(formula = 'y ~ a + I(a**2) + log(b) + c + d', data=data).fit()

where $y$ is a dependent variable depends on second degree polynomial of $a$, log of $b$ and linear in $c$ and $d$.

My question is in such a case what will my equation be?


If I understand your question, you're asking for the model form:

$$y = \alpha + \beta_0a + \beta_1a^2 + \beta_2\log(b) + \beta_3c + \beta_4d + \epsilon$$


  • $\epsilon \sim N(0,\sigma^2)$ is the OLS assumption
  • $a,b,c,d$ are in your data
  • $\beta_0,...\beta_4$ are your coefficients. $\alpha$ is the intercept which is added automatically when you use formulas, unless specified otherwise (by including a -1 in the formula).

See here for more on how Python formulas work.

| cite | improve this answer | |
  • $\begingroup$ Thank you @ilanman for you answer. Yes I am asking about the model form. $\endgroup$ – bikuser Jan 4 '17 at 11:45

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.