Giuseppe Ragusa

I have done a lot of work under the hood to provide an API for CovarianceMatrices. It needs some work. I might have some time next week to finish and...

I am always against clustering in the formula. Two reasons: the first is conceptual — for many models (OLS, panel, etc) the clustering is an after estimation process. The estimator...

The problem is there with `markdown` and not with latex literate file. The reason is that `%` does not initiate a comment in markdown.

I have changed the `output.call` to `comment = TRUE` which leave things as they were before this PR. When `comment=FALSE` the comment with the call output is not shown.

An argument for not automatically dropping missings can be made when the model is weighted. As of now ```julia lm(@formula(y~x), data=df, weights=aweights(df.w)) ``` with missing values in either `y`, `x`,...

Yes this is correct. On Sat, 16 Jul 2016 at 16:20, Benjamin Deonovic [email protected] wrote: > A few questions: D in your above example is the distribution of the >...

I agree that the an link invlink methods could suffice. However, as I said, I work on non standard problems. A reduction would be the following. The prior on a...

@brian-j-smith I think your option 2) is what I was thinking of, but a try/catch statement would probably be neater provided an parameter exception would exist. I will try to...

Thank you! I saw the extra three lines in master --- but I assume they were part of the new tagged version.

> Would that fix the issue I am having, which is that if rows of the data contains missing values, GLM discard those rows, but does not discard the corresponding...