Is there any method to append data to dataset?

[2young-2simple-sometimes-naive]

For example, if I would like to use under live trading, it would be tedious to dump_all, then build dataset again for each new input data. Is there anyway to directly append new data to existing dataset object, and generate features and labels?

[zhupr]

@2young-2simple-sometimes-naive Hi,
You can append data like this: python dump_bin.py dump_update --csv_path <new data dir> --qlib_dir <qlib_data_dir>.....

[2young-2simple-sometimes-naive]

@2young-2simple-sometimes-naive Hi,
You can append data like this: python dump_bin.py dump_update --csv_path <new data dir> --qlib_dir <qlib_data_dir>.....

Sorry, this is not what I meant. In the example workflow, data set is initialized by:
dataset = init_instance_by_config(task["dataset"])
Is there any way to append data to this dataset?

[you-n-g]

I guess your requirements are building the input dataset for recent data for high performance online inference.

The design of the underlying data infrastructure have considered such requirements.
Examples will be given in this PR #290 soon

[github-actions]

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

[jeffzhengye]

Also need this feature.