CODEWITHSUNDEEP
rregressionglm
ch-pub
ch-pub

Reputation: 1734

R: How to run regression (glm function) with some coefficients specified?

Suppose I have one dependent variable, and 4 independent variables. I suspect only 3 of the independent variables are significant, so I use the glm(y~ x1 + x2 + x3...) function. Then I get some coefficients for these variables. Now I want to run glm(y ~ x1 + x2 + x3 + x4), but I want to specify that the x1, x2, x3 coefficients remain the same. How could I accomplish this?

Thanks!

Upvotes: 1

Views: 1417

Answers (1)

Julián Urbano
Julián Urbano

Reputation: 8488

I don't think you can fit a model where some of the independent variables have fixed parameters. What you can do is create a new variable y2 that equals the predicted value of your first model with x1+x2+x3. Then, you can fit a second model y~y2+x4 to include it as an independent variable along with x4.

So basically, something like this:

m1 <- glm(y~x1+x2+x3...)
data$y2 <- predict(glm, newdata=data)
m2 <- glm(y~y2+x4...)

Upvotes: 1

Related Questions