search

ZhuZhouFan / TWMA

9 stars 5 forks source link

Request for Original Data to Reproduce Code #1

Open ftd0828 opened 1 month ago

ftd0828 commented 1 month ago

This is a very good piece of work! As a beginner in this field, I am eager to reproduce your code for learning purposes. Could you please provide the original data so that we can effectively reproduce the code? Thank you!

ZhuZhouFan commented 1 month ago

Sorry for the delayed response. As mentioned in our paper, the dataset used in our study was obtained from WIND, and unfortunately, I do not have the rights to make it publicly available. I recommend exploring alternative data sources that may be accessible for similar purposes. Thank you for your understanding.

zhikuichian commented 1 month ago

is it possible to provide a simple built module you did for us to do a infer evaluation?

ZhuZhouFan commented 1 month ago

@zhikuichian Regarding the evaluation, after using inference.py to obtain the estimated Triple-I weights, you can apply these weights to compute a weighted average of your specific signals and then construct a long-short portfolio. I suggest conducting portfolio backtesting using a public backtesting platform such as RiceQuant or JoinQuant, or utilizing open-source libraries like VNPY.