v0.3.0 — Segment-goal acquisition and segment PSD plotting

AutoSQUID v0.3.0 adds segment-yield acquisition for scan intervals where full clean traces are hard to collect.

• Added Config.segment_goal=(n_segments, segment_length).
• Segment mode keeps acquiring full n_points runs, then counts full clean windows from each usable prefix.
• Added goals.py for acquisition-goal logic.
• Added ledger fields for segment progress: goal_mode, goal_progress, segment_length, accepted_segments.
• Added backfill_ledger() for post-run ledger migration/backfill.
• Added plot_segment_psd_folder() for folder-based PSDs of usable segments.
• Added locate_usable=True so legacy jumped traces can be truncated before segment PSD plotting.
• Added optional segment_clean_only=True to drop failed individual segments from PSD overlays.

v0.2.0 — Clean-prefix acquisition and trace provenance

• Truncates jump/rail acquisitions to their clean pre-failure prefix.
• Separates ledger event, outcome, and accepted.
• Adds persistent post-hoc jump, rail, freeze, and dead-trace detection.
• Adds exact ledger-based resume counting.
• Adds usable-point filename tags and min_usable_frac.
• Adds plot_usable and clean-prefix analysis helpers.
• Automatically migrates pre-0.2 ledgers and saves the original as experiment_log.txt.bak.

v0.1.4 — Added configurable acquisition control

• Configurable DIFF/RSE/NRSE DAQ input mode.
• Optional temperature logging.
• Optional live jump checking.
• Improved auto-S-tune probing.
• Correct reset-attempt logging.

v0.1.3 — installable src layout + PyPI prep

• src/AutoSQUID/ + examples/ + docs/; pyproject (src layout) -> pip install -e .
• GPLv3 LICENSE
• export set_array_flux; TEMP_*.csv naming; jump_v=0.5; tol_V=20 mV
• publish metadata + Trusted-Publishing CI workflow