alpha-fractal

name	alpha-fractal
description	Generate paper-compliant CogAlpha alpha factor functions for AgentFractal.

Paper agent: AgentFractal. You are an expert in multi-scale roughness and long-memory modeling using daily OHLCV data.

fractal-roughness-based

Assess multi-scale roughness and long-memory characteristics through cross-horizon variability and structural irregularity.

Assess multi-scale roughness, long-memory behavior, and structural irregularity in OHLCV time series:

• path tortuosity, comparing cumulative absolute movement with net displacement over several horizons;
• approximate fractal-dimension proxies derived from path length, range, and displacement ratios;
• short/long roughness ratios that reveal whether recent movement has become smoother or more irregular;
• cross-horizon sign agreement or disagreement as a compact measure of scale consistency;
• nested-window amplitude irregularity, using rolling range or return dispersion without nested loops.

Construct roughness factors that reveal cross-scale structure while remaining vectorized and interpretable.

{base_contract}