Using Qlib Expressions with Custom 1-Minute OHLCV DataFrame (No .bin Files)

[AliMoeen]

❓ Questions and Help

Hello,

In my current HFT project I cannot use Qlib’s proprietary .bin file format.
I already have 1-minute OHLCV data in a pandas DataFrame (fetched from our backend), and I’d like to feed this directly into Qlib so that I can still use Qlib’s expressions (e.g., the ones in Alpha158).

I was hoping that DataHandlerLP.from_df could be an option, but it doesn’t seem to work with expressions.

Is there a straightforward way to supply my own DataFrame as the data source for Qlib (bypassing the .bin data directory), while still being able to use Qlib’s built-in expression engine?

[you-n-g]

We have some script like  https://github.com/microsoft/qlib/blob/main/scripts/dump_bin.py that can dump DataFrame-based data  directly into .bin files. Currently the expression engine is underlying the DataHander Design in a more infrastructure level up the raw data. So you can do that.

The raw data can have a other implementation rather than .bin file qlib/qlib/data/storage at main · microsoft/qlib. But curreently there is no other version of implementaion.