Answers are grounded in this paper and cite their sources.
Develops tests for functional weak white noise, suggesting improved inference accuracy.
The functional spherical autocorrelation function (fSACF), introduced by Yeh et al., provides a directional measure of serial dependence for functional time series, but existing inference is confined to an i.i.d. calibration. Building on this idea, we develop a class of portmanteau tests for functional weak white noise based on the fSACF. Unlike the conventional functional autocorrelation function, which is defined through squared norms and therefore captures only the magnitude of dependence, the fSACF preserves sign information in the Hilbert space and thus provides a richer characterization of serial dependence. We extend fSACF‐based inference from the i.i.d. setting to the more general weak white noise framework, which allows functional observations to be uncorrelated yet dependent. Under the null hypothesis, we derive the asymptotic distribution of the proposed test statistic and develop a block bootstrap procedure for critical value approximation and finite sample implementation. Simulation results show that the proposed procedure achieves accurate size control under weak white noise and is more robust than the Yeh test in non‐Gaussian settings. The practical usefulness of the method is illustrated through applications to electricity demand data and cumulative intradaily return curves.
No takes yet. Share an insight, caveat, or question.
Miao et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: