Setting intercept to 0 in lmBF?

rskarbez

Hi all,

I was wondering whether it is possible to force the intercept equal to 0 in lmBF. (Or, at least, I think that's what I need to do. Forgive my statistical shortcomings.)

What I have is a dataset in which I want to find the best fit from [linear, quadratic, cubic] fits, with the constraint that the intercept should be 0. I'd like to be able to say something like, "Assuming a polynomial of degree <= 3, and that the y-intercept is 0, the best fit for this data is f(x)."

Any help would be very much appreciated.

Best regards,
Rick

EJ

Hi Rick,

I'm not sure whether this can be accomplished (my personal probability is lower than 5%). If Richard does not reply, you could DM him; if you would you can post the answer here.

Cheers,
E.J.

rskarbez

Short answer: not currently possible

Richard's full response:

richarddmorey

mmm, no, the BayesFactor package won't allow that; the intercept has a noninformative prior placed on it that is analytically integrated out by the time the numerical integration algorithms even start. One would have to derive the posterior assuming 0 intercept. This could be done and tested in an hour or so by someone who knew what they were doing (but unfortunately I don't have time to do it).