THALis: phase-specific TDA

prep / execution / rest: where exactly am i cutting these? cue time isn't necessarily movement onset, and windows crossing the boundary could get messy

recorded extension (e7f5e25): the TDA pipeline adds multiphase analysis and generates another result set. The archive describes preparation, execution and rest as the phases of interest.

reference question: are phases defined by exported cue times, observed movement, or a fixed offset from trial onset? The answer determines what physiological interval each cloud represents.

Collector rest, cue, and imagery phases

comparison controls: keep the number of samples, preprocessing, embedding and filtration settings visible for each phase. Otherwise a phase difference could reflect cloud size, window length or amplitude changes rather than the hypothesized dynamics.

archived code references: e7f5e25.

• questions to recover from the run: how are transitions handled, are windows allowed to cross phase boundaries, and are repeated trials compared within a session or pooled? The commit summary does not provide a significance test or generalizable phase-separation result; retain this as an analysis extension and a pointer to its artifacts.