Overview

Axis-agnostic metrics that operate on any (date, value) series produced by an upstream cell metric — information coefficient (IC) time series, \(\beta\) time series, CAAR time series, an external factor return, etc. They do not care which cell produced the series.

Not the same as Mode.TIMESERIES

Mode.TIMESERIES is the dispatch regime for n_assets == 1 (set on FactorProfile.mode); the metrics on this page run on a (date, value) series regardless of which mode produced it.

Metric	Role	Page
Out-of-sample decay across multiple splits	Robustness	oos
Theil-Sen monotonic trend with augmented Dickey-Fuller (ADF) persistence flag	Diagnostic	trend
Sign-consistency / hit rate of the series	Diagnostic	hit_rate

Use these against the IC series from ic.compute_ic, the CAAR series from caar.compute_caar, or any externally constructed (date, value) DataFrame to layer decay, drift, and sign-stability checks on top of the cell's primary inferential test.