0

the irls R function carries out Iteratively Re-weighted Least Squares algorithm (Source: https://www.rdocumentation.org/packages/msme/versions/0.5.3/topics/irls).

One of the argument that can be passed in the irls function is the offset. According to the website above, offset is defined as:

"this can be used to specify an a priori known component to be included in the linear predictor during fitting. This should be 0 or a numeric vector of length equal to the number of cases"

Which I am not so sure what it means (could be something really simple, but I am not sure).

What is the argument offset for in this R function?

Thank you,

jschnieder
  • 213
  • 2
  • 4
  • Where "there" is the parameter? – Tim Jun 29 '20 at 19:59
  • the functions for calculating the posterior mode of a Bayesian logistic regression model, and for the posterior variance of a Bayesian logistic regression model – jschnieder Jun 29 '20 at 20:00
  • these functions calculate posterior mode and variance, in order to enable the Metropolized iid sampling. – jschnieder Jun 29 '20 at 20:01
  • I think the functions tries to calculate posterior mode and variance by applying the Iteratively Reweighted Least Squares – jschnieder Jun 29 '20 at 20:03
  • Unless you give us more details on what you are referring to, it’s not clear what you mean. Logistic regression does not have “offset” and offset in regression is constant, so is not updated https://stats.stackexchange.com/questions/11182/when-to-use-an-offset-in-a-poisson-regression – Tim Jun 29 '20 at 21:52
  • Thanks for edit, it’s the offset from above link that seems to answer your question. – Tim Jun 29 '20 at 22:02

0 Answers0