You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Apologies if this is the wrong way to reach out for this topic, but I am curious on the details of Hansen's Skewed T distribution as you've implemented in the arch package. I've read his paper and some other papers in the literature, but I have not found a source that derives the PPF, CDF, and moments for the distribution.
I am working on a python package myself, and I want to include an implementation of Hansen's Skewed T. I could try to derive the CDF / PPF myself or use numerical methods (and measure the skewness / kurtosis with Monte Carlo), but I prefer a closed form expression with a reliable source if possible. I am not an academic, so I feel much more comfortable programming existing distributions than trying to derive or add to the existing math.
Obviously, you have these methods implemented, and I have seen a couple of other implementations of the Skewed T distribution floating around complete with PPF's, CDF's etc, so the derivation must be out there somewhere. If you could refer me to the paper that does this work that would be greatly appreciated!
Thanks for the good work--the Arch package is one of my favorites in the Python ecosystem, one of the few that I use regularly besides the basics (Numpy, Pandas, Scipy, Matplotlib, etc)!
(EDIT: Oops, wrote this from my work account.)
The text was updated successfully, but these errors were encountered:
Apologies if this is the wrong way to reach out for this topic, but I am curious on the details of Hansen's Skewed T distribution as you've implemented in the arch package. I've read his paper and some other papers in the literature, but I have not found a source that derives the PPF, CDF, and moments for the distribution.
I am working on a python package myself, and I want to include an implementation of Hansen's Skewed T. I could try to derive the CDF / PPF myself or use numerical methods (and measure the skewness / kurtosis with Monte Carlo), but I prefer a closed form expression with a reliable source if possible. I am not an academic, so I feel much more comfortable programming existing distributions than trying to derive or add to the existing math.
Obviously, you have these methods implemented, and I have seen a couple of other implementations of the Skewed T distribution floating around complete with PPF's, CDF's etc, so the derivation must be out there somewhere. If you could refer me to the paper that does this work that would be greatly appreciated!
Thanks for the good work--the Arch package is one of my favorites in the Python ecosystem, one of the few that I use regularly besides the basics (Numpy, Pandas, Scipy, Matplotlib, etc)!
(EDIT: Oops, wrote this from my work account.)
The text was updated successfully, but these errors were encountered: