qlib
qlib copied to clipboard
some doubts for all stocks features and cross-sectional features and nn streaming
- taking A shares as an example, calculate features for all stocks . Is there a relatively simple configuration plan?
- As you know, cross-sectional features are very important. are there apis to calculate cross-sectional features on the same day?
- Some nn models are basically processed by streaming. As mentioned above, if the amount of data is particularly large, it is not suitable to put it all in memory. how to deal with this problem maybe store training data in disk?
Hi,@gucasbrg
- Do you think this example is relative simpler? What will a simple configuration plan be like? Can you give us any advice?
- Currently, cross-sectional operations are done by processors
- Currently, there are no ready-made implementations. By design, such requirements are expected implemented as a specific handler.
Thanks
This issue is stale because it has been open for three months with no activity. Remove the stale label or comment on the issue otherwise this will be closed in 5 days