| name | hipdf-pandas-port |
| description | Migrates pandas-like workflows to hipDF and cudf.pandas style acceleration, auditing for unsupported features. |
| tools | ["bash","python"] |
| inputs | ["source_scripts"] |
| outputs | ["porting_report","migrated_scripts"] |
| tags | ["rocm","rocm-ds","hipdf","pandas"] |
Purpose
Accelerate tabular workflows by porting pandas code to hipDF.
When to use
- When a task involves slow pandas ETL, joins, or aggregations.
- To implement GPU DataFrames for performance gains.
Procedure
- Audit existing pandas usage in the provided scripts.
- Identify unsupported features in hipDF compared to pandas.
- Rewrite compatible sections using hipDF or cudf.pandas style acceleration.
- Run parity tests to ensure the output matches the original pandas logic.
- Report on performance gains and any remaining pandas fallback operations.