AI RESEARCH
Energy Consumption of Dataframe Libraries for End-to-End Deep Learning Pipelines:A Comparative Analysis
arXiv CS.AI
•
ArXi:2511.08644v3 Announce Type: replace-cross This paper presents a detailed comparative analysis of the performance of three major Python data manipulation libraries - Pandas, Polars, and Dask - specifically when embedded within complete deep learning (DL)