Forum: help

General linear regression with log-linear [ Reply ]
By: Vincent Goffredo on 2019-10-12 07:27

Good evening everyone,

I state that I am not very experienced in R or other statistical software.

I have a problem, which from the inexperienced of the sector, appears to me more than complicated.

I have two columns of X (estimated) and Y (real) values related to each other (according to the Spearman index)

I have to use a Generalized Linear Regression with Log-Linear model and random component with Binomial Negative.

the equation I have to get is like:

Y = exp(b0) * exp(b1*du) * X^b2

where du is a simple dummy variable that is always 0 except a case that holds 1.

how do I estimate the parameters b?