Please note, per the posting guide linked below:
"*Questions about statistics:* The R mailing lists are primarily intended
for questions and discussion about the R software. However, questions about
statistical methodology are sometimes posted. If the question is well-asked
and of interest to some
Hello,
I'm currently reviewing how to correctly implement `glmnet` and am having a
hard time understanding why the results seem to be different between each
method when `intercept = TRUE/FALSE` as I thought it should just drop the
intercept from the model. However, it seems to be acting a bit d
You can't. But in order to be able to predict for states that _were_ in the
training data, the coefficient cannot be NA.
On February 23, 2021 8:40:43 AM PST, b...@denney.ws wrote:
>How should you be able to make a prediction (using this type of model)
>from a state where there is no data such as
How should you be able to make a prediction (using this type of model) from a
state where there is no data such as treatment="C" in my example?
-Original Message-
From: Jeff Newmiller
Sent: Tuesday, February 23, 2021 11:10 AM
To: r-help@r-project.org; b...@denney.ws
Subject: Re: [R] pri
Model equations do not normally have conditional forms dependent on whether
specific coefficients are NA or not. If you assign NA to a coefficient then you
will not be able to predict outputs for input cases that you should be able to.
Zero allows these expected cases to work... NA would prevent
Hello,
I'm working on a survreg model where the full data are subset for modeling
individual parts of the data separately. When subsetting, the fit variable
("treatment" in the example below) has levels that are not in the data. A
work-around for this is to drop the levels, but it seems inacc
6 matches
Mail list logo
