Investigating Differences Between RSI Feature Expressions

Summary

This forum post raises a feature-engineering question: why a named 14-period RSI feature, a function-based RSI expression, and a manually written RSI formula return different values. The user describes extracting all three through default basic-feature and derived-feature modules for one Chinese stock over a specified date range, then asks whether the discrepancy is expected or reflects a misunderstanding of RSI.

The post provides the feature names, a rolling gain-versus-absolute-change formula, and enough extraction context to frame a reproducible comparison. It does not include a reply or resolve the discrepancy. Possible issues such as differing RSI conventions, initialization, warm-up history, or feature definitions are not analyzed in the document, so it cannot establish which calculation is correct. Its value is as a concrete diagnostic case for researchers comparing technical indicator implementations across data pipelines.

Key ideas

• The user compares a named RSI feature, a function-based expression, and a manually expressed formula.
• The values were extracted through two feature modules using their default settings.
• The example specifies a single Chinese equity and a date interval for the comparison.
• The document contains no answer, so the source of the differences remains unresolved.

Tags

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