ALGORITHM FOR OPTIMAL TOOL SELECTION FOR DATA ANALYSIS IN PYTHON IN THE E-COMMERCE ANALYTICS INDUSTRY

Shkokov I. O.

Аrticle addresses the problem of selecting the optimal data analysis tool for e-commerce web analytics. A comparative performance study of four popular Python libraries—Pandas, Polars, DuckDB, and PySpark—is conducted using typical business queries. Based on the experimental results measuring data processing time, the strengths of each library are identified according to task complexity and data volume. As a result, a decision tree algorithm is proposed to help data specialists choose the fastest tool for a specific analytical task, thereby significantly increasing workflow efficiency.

Keywords:data analysis, Python, optimization, web analytics, e-commerce, data libraries, performance.