i have an observable which is a ratio of two others that are correlated (where the correlation is changing over time). Therefore a t-distribution is a much better assumption for the residuals than the normal distribution. How can I implement this in D2D?
Hi,
i have an observable which is a ratio of two others that are correlated (where the correlation is changing over time). Therefore a t-distribution is a much better assumption for the residuals than the normal distribution. How can I implement this in D2D?
best,
Lorenz