Work

Forecastability Profiles: Horizon-Resolved Predictive Information under Logarithmic Loss — Replication Code (v1.0.0)

Year: 2026

Type: software

Abstract: Python replication code for the manuscript "Forecastability Profiles: Horizon-Resolved Predictive Information under Logarithmic Loss" by Peter M. Catt. The archive contains the source code, dependency list, and run configuration required to reproduce the empirical illustrations reported in the manuscript, including the synthetic AR(1), noisy logistic map, pure seasonal autoregressive process, monthly sunspot counts, and S&P 500 daily log-return examples. The code reproduces the forecastability profiles, AR(1) analytical comparison, and profile-level block-surrogate integrated-mass test described in the paper.

Source: Zenodo (CERN European Organization for Nuclear Research)

DOI: 10.5281/zenodo.19931874

Author: Catt, Peter Maurice

Language: English

Topic: Complex Systems and Time Series Analysis

Subfield: Economics and Econometrics

Field: Economics, Econometrics and Finance

Domain: Social Sciences

Open Access status: green

OpenAlex ID: W7159574149

Cites:

Cited by:

Related to: 10