Add path_to_csv arg to gp_fromdata.py and bo_fromdata.py

[JasonBernstein1]

In the gp_fromdata.py and bo_fromdata.py scripts, datasets are loaded with the data_processing.load_data function. This function has a path_to_csv arg, but it is not exposed in these scripts. We should add this arg as a command-line flag and let it be set in this script through this function.

This will make the scripts more robust to the location of where the scripts are called.

[JasonBernstein1]

Resolved in commit 10708e4 by requiring explicit dataset registration in DATASET_CONFIG rather than exposing --path_to_csv. This ensures proper validation and makes supported datasets discoverable via --help. Added hst_H as an example.