Niko Huurre

How about adding [matrix normal distribution](https://en.wikipedia.org/wiki/Matrix_normal_distribution)? Row-wise and column-wise multi-normal distributions are special cases of it. ```stan Y ~ matrix_normal(Mu, Sigma, identity_matrix(Ncols)); Y ~ matrix_normal(Mu, identity_matrix(Nrows), Sigma); ```

The expected change is adding `if (size(y)==0)` branches in `stan/math/prim/fun/unit_vector_constrain.hpp` and a test case in `test/unit/math/prim/fun/unit_vector_constrain.hpp`. However, I think the issue is wrong. As currently implemented, the constraining transform does...

AFAIK sparse matrices are most useful when dense representation would inefficient. Automatically converting to a dense matrix seems like a footgun to me. What's the purpose of supporting such "sparsity...

> The problem is that when you apply exp() to a sparse matrix, the result is dense. And my question is: if applying exp() to a sparse matrix defeats the...

> Another alternative we should consider is to simply not allow functions where f(0) != 0 to apply elementwise to sparse matrices. Yes, come to think of it, that would...

> what do we want to do for `sparse_matrix[...]` A dense return type is definitely wrong here! The constraining transform must apply only to nonzero elements. `lb_constrain()` is not `exp()`,...

> @ethan-alt: The definition of a mass function for a random variable X evaluated at value x is "the probability that X is equal to x, i.e., `p(x) = P(X...

@betanalpha > The problem here isn't with the CmdStan argument parser but rather with that recently introduce feature of trying to modify the CmdStan configuration based on the given Stan...

> There are many zero-dimensional spaces and hence many different ways to extrapolate to the N = 0 boundary case of R^N. There are indeed many zero dimensional spaces (although,...

> will the comment header report "adaptation=False"? It says `adapt engaged=1` so it still reports technically incorrect config. However, adaptation is irrelevant for zero-dimensional models so only possible problem would...