Migrating StockRanker Training Data from BigQuant 1.0 to 3.0

Summary

This guide explains how to carry training data for a StockRanker model from BigQuant’s 1.0 environment into version 3.0. In the older environment, it exports the training dataset with date, instrument, label, and the selected feature columns. The file is then placed in the newer environment and written as a temporary data source for use by the StockRanker module.

For compatibility, the guide says to set the module’s data sort field to date. It reports that feature importance is consistent between the two environments after migration. That comparison is the only evidence offered: the document gives no performance metrics, validation results, or details about whether predictions remain identical. The steps describe data transfer and configuration, rather than a general model portability guarantee, so users should verify behavior with their own model and dataset.

Key ideas

• Export the training data with date, instrument, label, and the model’s selected feature columns.
• Import the exported file into version 3.0 and register it as a temporary data source.
• Set the StockRanker data sort field to date to support compatibility with the older environment.
• The guide reports matching feature importance across versions, but provides no performance validation.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.