bgoodri
bgoodri
After putting in stuff to deal with the diagonal, we still have the problem that the optimal value of the standard deviation is zero numerically (-1 million on the log...
I have a unpushed branch that does this for the Bessel K function only.
The particularity of Stan that we are dealing with here is that if we allow the order to be a real number, then we have to be able to differentiate...
Differentiating with respect to x is not so bad. Differentiating with respect to nu is harder. We would need to implement (with numerical accuracy) things like http://functions.wolfram.com/Bessel-TypeFunctions/BesselI/20/01/01/ On Tue, Jun...
We skip it now but only because we restrict nu to be an integer. On Tue, Jun 28, 2016 at 3:45 AM, Pascal PERNOT [email protected] wrote: > OK, I see...
Yeah, I would just define a function in the Stan language real my_bessel(int nu4, real x) { real nu; nu That would still hold if we constrain nu to be...
This sounds good. I'll test it on Linux. Thanks!
It is a relatively new R development. We'll deal with it. On Sun, Feb 7, 2021 at 2:25 PM jpkrooney wrote: > Also experiencing this since updating to Big Sur...
We should fix ADVI in Stan so that it does not report values greater than the largest floating point number. On Tue, May 9, 2017 at 5:22 PM, Daniel Lee...
Yes, and not that I know of. On Wed, May 10, 2017 at 2:50 PM, Bob Carpenter wrote: > @bgoodri: is your proposal todo censoring and report the max double...