fracdiff DOC: Jupyter notebook showing how to optimize "d" and "window" params

Apr 13 '23 23:04 replacementAI

Lmk if any changes need to be made @simaki. Thanks for taking a look 😃

Apr 13 '23 23:04 replacementAI

Btw this is how it looks like https://github.com/replacementAI/fracdiff/blob/main/examples/Frac%20Diff%20Example.ipynb

Apr 14 '23 18:04 replacementAI

I think in block 7, I need to elaborate that the frac diff model outperforms the 3 other models. I also want to reduce the search space window to 147 (21 days per month * 7 hours per trading day).

Apr 14 '23 18:04 replacementAI

I still need some time to work on this, Ive been a bit busy

May 07 '23 17:05 replacementAI

@simaki, sorry for the late commit, I've finished up the example. The target is the log return of a L/S portfolio where ETH is long and BTC is short. I picked crypto because it might have more alpha than equities, and I also constructed the portfolio this way in order to compensate for the market factor, which I believe bitcoin is a good proxy for.

Jul 14 '23 22:07 replacementAI

All thats left is to figure out where to put this jupyter notebook. Its already in the examples folder, but should I just put a hyperlink on the readme? Or should I put the hyperlink, and some pictures from the notebook?

Jul 14 '23 22:07 replacementAI