Complete configuration reference

All 189 live configuration keys are listed here with their installed defaults and allowed values. This page is generated from Workbench's arguments.CONFIG_ARGUMENTS through describe_config, the same definitions used by the scientific normaliser.

A partial config or settings object may contain only the keys you change. Configurable registered actions store the complete resolved settings in provenance.effective_config, alongside the changed-settings summary; actions without configuration use None. Explicit action arguments and ordered preprocessing remain distinct from that settings mapping. Defaults come from the installed package, never saved run history. See settings precedence and compatibility loading.

Selection, core period, and actogram

Key	Type	Default	Allowed values	Units	Constraints	What it changes
start	str or null	null	—	ISO date or datetime		Inclusive start of the analysis interval; null starts at the first sample.
end	str or null	null	—	ISO date or datetime		Inclusive end of the analysis interval; null ends at the last sample.
exclusions	list	[]	—	dated intervals		Intervals treated as unavailable without altering the source recording.
analysis_channel	str	"activity"	—	named choice or label		Measurement key analysed; each channel keeps its declared sum or mean aggregation.
period_hours	float	24.0	—	hours	≥ 1; ≤ 240	Reference cycle used for folding, cosinor, and circular quantities.
period_min_hours	float	18.0	—	hours	≥ 0; less_than:period_max_hours; positive	Shortest candidate period in spectral and instantaneous-phase searches.
period_max_hours	float	30.0	—	hours	≥ 0; greater_than:period_min_hours; positive	Longest candidate period in spectral and instantaneous-phase searches.
bin_minutes	int	15	—	minutes	≥ 1; ≤ 180	Width of the regular analysis grid in minutes.
double_plot	bool	true	—	boolean switch		Show each actogram row beside the following cycle.
actogram_start_hours	float	0.0	—	clock hours		Clock time at the left edge of an actogram row; wraps within one day at minute resolution.
actogram_mode	str	"scaled"	scaled, normalized, threshold, percentile, even; aliases: quantile → scaled	named choice or label		Activity marks: scaled uses the record's 99th percentile; normalized scales each day; threshold is binary; percentile uses quantile levels; even uses equal-width levels. The old unsupported label quantile retains its former scaled output; request percentile explicitly for quantile levels.
actogram_threshold	float	1.0	—	recording units		Minimum activity drawn as active in threshold actograms; ignored by other display modes.
actogram_quantiles	int	5	—	count	≥ 2; ≤ 10	Number of activity levels in percentile or even actograms; does not alter numerical analysis.
actogram_y_scale	str	"fit"	fit, fixed	named choice or label		fit displays every day together; fixed pages through the record at a constant row height.
actogram_page_days	int	14	—	days	≥ 1; ≤ 365	Day rows per page when the actogram Y scale is fixed; does not select analysis days.
cursor_start_hours	float	6.0	—	hours		Start clock hour of the activity-profile interval selected by the two cursors.
cursor_end_hours	float	18.0	—	hours		End clock hour of the activity-profile interval selected by the two cursors.
bout_threshold	float	1.0	—	recording units		Activity threshold used to identify active samples for bout scoring.
bout_min_minutes	int	10	—	minutes	≥ 1	Shortest active bout retained after permitted gaps are joined.
bout_max_gap_minutes	int	10	—	minutes	≥ 0	Longest inactive gap joined inside an activity bout; zero keeps every gap.
onset_off_hours	float	6.0	—	hours	≥ 0.25; ≤ 24	Required inactive window preceding an activity onset or following an offset.
onset_on_hours	float	6.0	—	hours	≥ 0.25; ≤ 24	Required active window following an activity onset or preceding an offset.
onset_threshold_percent	float	80.0	—	percent	≥ 1; ≤ 99	Percentage of nonzero activity counts treated as active for daily onset and offset detection.
onset_overrides	dict	{"onset":{},"offset":{}}	—	clock hours by date		Explicit per-date onset or offset replacements retained in provenance.
show_marker_paths	bool	false	—	boolean switch		Connect daily onset and offset markers on the actogram without changing the measured times.
show_marker_fits	bool	false	—	boolean switch		Display fitted onset and offset lines on the actogram without changing their estimates.
fit1_start	str or null	null	—	ISO date or datetime		First included date in the first daily-marker fit, normally before a perturbation.
fit1_end	str or null	null	—	ISO date or datetime		Last included date in the first daily-marker fit.
fit2_start	str or null	null	—	ISO date or datetime		First included date in the second daily-marker fit, normally after a perturbation.
fit2_end	str or null	null	—	ISO date or datetime		Last included date in the second daily-marker fit.
phase_shift_day	str or null	null	—	ISO date or datetime		Date at which the two daily-marker fits are compared to measure a phase shift.
periodogram_alpha	float	0.01	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for periodogram; smaller values require stronger evidence.
periodogram_method	str	"lomb"	lomb, chi_square, f	named choice or label		Periodogram supplying the rhythmicity verdict; distinct from the estimator selected by period_method.

Extended period estimation

Key	Type	Default	Allowed values	Units	Constraints	What it changes
period_method	str	"lomb"	lomb, chi_square, f, fft_nlls, mesa, mfourfit, spectrum_resampling, jtk, ejtk	named choice or label		Single period estimator requested by estimate_period; independent of the periodogram used for a rhythmicity verdict.
period_methods	list	["lomb"]	lomb, chi_square, f, fft_nlls, mesa, mfourfit, spectrum_resampling, jtk, ejtk	dimensionless		Estimators requested by compare_periods; the default runs Lomb-Scargle only.
phase_reference	str	"zero"	zero, window	named choice or label		Origin used when a period estimator reports phase.
phase_units	str	"hours"	hours, circadian	named choice or label		Report phase in clock hours, or rescale it so one fitted period spans 24 circadian hours.
period_detrend	str	"none"	amp_baseline, asymmetric_least_squares, first_difference, frequency, kernel, linear, lowess, moving_median, none, polynomial, robust_linear, running_mean, savitzky_golay; aliases: robust → robust_linear, huber → robust_linear, first-difference → first_difference, difference → first_difference, diff → first_difference, cubic → polynomial, bicubic → polynomial, poly3 → polynomial, poly6 → polynomial, degree6 → polynomial, baseline → kernel, moving_average → running_mean, running_median → moving_median, median → moving_median, loess → lowess, savitzky-golay → savitzky_golay, savgol → savitzky_golay, amp&baseline → amp_baseline, asymmetric-least-squares → asymmetric_least_squares, asls → asymmetric_least_squares, als → asymmetric_least_squares	named choice or label		Baseline-removal method applied before extended period estimation.
period_detrend_window_hours	float	24.0	—	hours	≥ 1; ≤ 240	Biological-time window for local baseline removal before period estimation; not a period search bound.
period_detrend_polynomial_degree	int	3	—	count	≥ 0	Polynomial degree used when period detrending is polynomial.
period_detrend_min_valid_fraction	float	0.5	—	dimensionless	≥ 0; ≤ 1	Minimum valid fraction required inside a detrending window.
period_detrend_bandwidth_hours	float or null	null	—	hours	≥ 1e-06	Kernel bandwidth used by kernel baseline detrending.
period_detrend_low_cut_hours	float	45.0	—	hours	≥ 1e-06; greater_than:period_detrend_high_cut_hours	Longest period retained by frequency-filter detrending.
period_detrend_high_cut_hours	float	4.0	—	hours	≥ 1e-06; less_than:period_detrend_low_cut_hours	Shortest period retained by frequency-filter detrending.
period_detrend_filter_order	int	2	—	count	≥ 1	Digital-filter order used by frequency-filter detrending.
period_detrend_lowess_fraction	float or null	null	—	dimensionless	≥ 1e-06; ≤ 1	Finite-record fraction used by each LOWESS local fit; null derives it from the detrending window.
period_detrend_lowess_iterations	int	3	—	count	≥ 0; ≤ 20	Robust residual-reweighting passes used by LOWESS detrending.
period_detrend_asls_smoothness	float	1000000.0	—	dimensionless	≥ 1e-12	Second-difference smoothness penalty used by asymmetric least-squares detrending.
period_detrend_asls_asymmetry	float	0.01	—	dimensionless	≥ 1e-06; ≤ 0.499999	Weight given to points above an asymmetric least-squares baseline.
period_detrend_asls_iterations	int	10	—	count	≥ 1; ≤ 100	Reweighting passes used by asymmetric least-squares detrending.
nlls_max_components	int	5	—	count	≥ 1; ≤ 25	Maximum cosine components considered by the fast Fourier transform nonlinear least-squares estimator.
nlls_circadian_min	float or null	null	—	hours	≥ 0; less_than:nlls_circadian_max|period_max_hours; positive	Legacy name for the FFT-NLLS component-selection lower bound, in hours. Null follows period_min_hours, including ultradian searches.
nlls_circadian_max	float or null	null	—	hours	≥ 0; greater_than:nlls_circadian_min|period_min_hours; positive	Legacy name for the FFT-NLLS component-selection upper bound, in hours. Null follows period_max_hours, including infradian searches.
nlls_improvement_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for nlls improvement; smaller values require stronger evidence.
mesa_model_length	int or null	null	—	dimensionless	≥ 2	Autoregressive model length for maximum entropy spectral analysis; null derives it from sampling frequency and available data.
chi_square_spline_interval_hours	float or null	null	—	hours	≥ 0.01; ≤ 6	Optional interpolation grid before chi-square and F periodograms; null retains the original binned grid.
chi_square_normalisation	str	"sokolove_bushell"	legacy, sokolove_bushell	named choice or label		sokolove_bushell is the published chi-square statistic; legacy divides it by cycle count to reproduce releases through 0.5.0.
mfourfit_harmonics	int	4	—	count	≥ 0; ≤ 4	Additional harmonics above the main cosine in the mFourFit waveform model.
mfourfit_step_hours	float	0.05	—	hours	≥ 0.001; ≤ 1	Spacing between candidate periods in the mFourFit grid search.
sr_iterations	int	1000	—	count	≥ 10; ≤ 100000	Bootstrap draws averaged by spectrum resampling; changing this changes the estimator, not only its speed.
sr_bandwidth	float or null	null	—	cycles per hour		Spectral smoothing bandwidth for spectrum resampling; null chooses it by cross-validation.
sr_trim_high_frequency	bool	true	—	boolean switch		Remove high-frequency spectral noise before spectrum-resampling bootstrap draws.
sr_grid_points	int	256	—	count	≥ 8; ≤ 4096	Frequencies interpolated into the spectrum-resampling search band; determines peak resolution.
sr_seed	int	20260901	—	integer seed		Fixed random-generator seed for sr; retain it to reproduce the stochastic calculation.

JTK_CYCLE and empirical JTK_CYCLE

Key	Type	Default	Allowed values	Units	Constraints	What it changes
jtk_periods	list	[20.0,24.0,28.0]	—	hours		Template periods tested by JTK_CYCLE and empirical JTK_CYCLE.
jtk_asymmetries	list or null	null	—	dimensionless	≥ 0.05; ≤ 0.95	Optional active:inactive waveform ratios for JTK templates; null uses symmetric templates.
jtk_phase_step_hours	float or null	null	—	hours	≥ 0.01; ≤ 24	Spacing of candidate phases for rank-based rhythm templates; null uses the method's sampling-based step.
ejtk_permutations	int	1000	—	count	≥ 10; ≤ 100000	Random permutations used for the empirical rank-based rhythmicity null distribution.
jtk_correction	str	"bh"	bh, bonferroni, sidak, none	named choice or label		Multiple-testing correction within the jtk comparison family, not across unrelated analyses.
jtk_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for jtk; smaller values require stronger evidence.
jtk_seed	int	20260901	—	integer seed		Fixed random-generator seed for jtk; retain it to reproduce the stochastic calculation.
jtk_max_points	int	96	—	count	≥ 4; ≤ 1024	Maximum time points passed to rank-based rhythmicity tests; denser recordings are binned first.

Circular and instantaneous analyses

Key	Type	Default	Allowed values	Units	Constraints	What it changes
circular_test	str	"permutation"	permutation, watson_williams	named choice or label		permutation compares phases by random relabelling; watson_williams requires sufficient within-group phase concentration.
circular_permutations	int	2000	—	count	≥ 10; ≤ 100000	Label permutations used to test between-group phase differences.
circular_min_resultant	float	0.7	—	dimensionless	≥ 0; ≤ 1	Minimum phase concentration before the parametric circular test is considered.
circular_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for circular; smaller values require stronger evidence.
circular_seed	int	0	—	integer seed		Fixed random-generator seed for circular; retain it to reproduce the stochastic calculation.
circular_correction	str	"sidak"	bh, bonferroni, sidak, none	named choice or label		Multiple-testing correction within the circular comparison family, not across unrelated analyses.
instantaneous_band_min_hours	float or null	null	—	hours	less_than:instantaneous_band_max_hours|period_max_hours; positive	Shortest period retained by the instantaneous-phase bandpass; null follows period_min_hours.
instantaneous_band_max_hours	float or null	null	—	hours	greater_than:instantaneous_band_min_hours|period_min_hours; positive	Longest period retained by the instantaneous-phase bandpass; null follows period_max_hours.
instantaneous_edge_trim_cycles	float	1.5	—	cycles	≥ 0.25; ≤ 5	Filter-edge duration removed from each end, measured in cycles.

Phase response and longitudinal analyses

Key	Type	Default	Allowed values	Units	Constraints	What it changes
prc_reference	str	"nocturnal"	nocturnal, diurnal	named choice or label		Circadian-time convention for stimulus timing: nocturnal activity onset is at circadian time 12, diurnal onset at 0.
prc_marker	str	"onset"	onset, peak	named choice or label		Daily marker used to score a phase response: activity onset or waveform peak.
prc_min_onsets	int	5	—	count	≥ 2; ≤ 365	Usable daily markers required in the pre-stimulus fit before scoring an animal's response.
prc_min_animals	int	8	—	count	≥ 3; ≤ 1000	Independent animals required before fitting a population phase response curve.
prc_harmonics	int	1	—	count	≥ 1; ≤ 2	Harmonics in the circular regression fitted to the phase response curve.
prc_dead_zone_hours	float	0.5	—	hours	≥ 0; ≤ 12	Maximum absolute fitted phase shift classified as a response-curve dead zone.
prc_comparison_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for prc comparison; smaller values require stronger evidence.
prc_comparison_correction	str	"sidak"	sidak, bonferroni, none	named choice or label		Multiple-testing correction within the prc comparison comparison family, not across unrelated analyses.
longitudinal_min_subjects	int	3	—	count	≥ 3; ≤ 1000	Subjects with complete repeated measurements required for a longitudinal comparison.
longitudinal_min_epochs	int	3	—	count	≥ 3; ≤ 1000	Ordered visits required for a longitudinal comparison; two visits belong to paired measures.
longitudinal_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for longitudinal; smaller values require stronger evidence.
longitudinal_correction	str	"sidak"	bh, bonferroni, sidak, none	named choice or label		Multiple-testing correction within the longitudinal comparison family, not across unrelated analyses.
longitudinal_sphericity	str	"greenhouse_geisser"	greenhouse_geisser, huynh_feldt, none	named choice or label		Correction to repeated-measures degrees of freedom: Greenhouse-Geisser, Huynh-Feldt, or none.

Splitting and dose response

Key	Type	Default	Allowed values	Units	Constraints	What it changes
splitting_min_days	float	10.0	—	days	≥ 1	Recording duration required before testing for two circadian components; spectral resolution still depends on total duration.
splitting_min_separation_hours	float	0.5	—	hours	≥ 0	Absolute floor on separation of two candidate periods, in addition to the record's resolution limit.
splitting_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for splitting; smaller values require stronger evidence.
splitting_hold_days	int	3	—	days	≥ 1	Consecutive windows required to support persistent splitting rather than an isolated second peak.
splitting_harmonic_tolerance	float	1.0	—	resolution elements	≥ 0	Distance, in spectral resolution elements, inside which two peaks cannot be distinguished from harmonics.
dose_min_levels	int	4	—	count	≥ 4	Distinct stimulus doses required before fitting the three-parameter Hill response curve.
dose_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for dose; smaller values require stronger evidence.
dose_ct_tolerance_hours	float	2.0	—	hours	≥ 0	Allowed spread of stimulus circadian times within a dose series; prevents mixing timing and dose effects.
dose_bootstrap	int	2000	—	count	≥ 50; ≤ 20000	Animal-level bootstrap resamples used for dose-response uncertainty.

Entrainment, masking, feeding, and sleep

Key	Type	Default	Allowed values	Units	Constraints	What it changes
entrainment_reference	str	"cycle_off"	cycle_off, cycle_on; aliases: lights_on → cycle_on, lights_off → cycle_off	named choice or label		Daily cue transition used for phase angle: cycle_on or cycle_off. lights_on and lights_off are retained aliases.
entrainment_zeitgeber	str	"light"	—	named choice or label		Declared cue kind used for entrainment, such as light or food; must match the experiment's schedule.
entrainment_marker	str	"onset"	onset, offset, acrophase	named choice or label		Daily onset, offset or acrophase used to measure angle to the chosen cue transition.
entrainment_min_days	int	5	—	days	≥ 3; ≤ 365	Usable scheduled days required before reporting an entrainment phase angle.
entrainment_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for entrainment; smaller values require stronger evidence.
reentrainment_criterion_sd	float	2.0	—	baseline standard deviations	≥ 0.5; ≤ 10	Tolerance around the new steady-state angle in multiples of the animal's pre-shift day-to-day scatter.
reentrainment_hold_days	int	3	—	days	≥ 1; ≤ 30	Consecutive days inside the tolerance required before declaring re-entrainment.
reentrainment_min_days_each_side	int	4	—	days	≥ 2; ≤ 60	Usable days required on each side of a schedule shift to establish baseline and endpoint.
masking_control_days	int	7	—	days	≥ 2; ≤ 60	Days before a pulse supplying activity from the same clock-time control window.
masking_min_control_days	int	3	—	days	≥ 2; ≤ 60	Usable pre-pulse control days required to compare acute activity with baseline.
masking_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for masking; smaller values require stronger evidence.
faa_window_hours	float	3.0	—	hours	≥ 0.25; ≤ 12	Window before food becomes available in which anticipatory activity is measured.
faa_hold_days	int	3	—	days	≥ 1; ≤ 30	Consecutive days above baseline criterion required for food anticipation.
faa_criterion_sd	float	2.0	—	baseline standard deviations	≥ 0; ≤ 10	Food-anticipation criterion in multiples of the animal's pre-schedule standard deviation.
faa_min_days	int	5	—	days	≥ 2; ≤ 60	Pre-schedule baseline days required to establish a food-anticipation criterion.
faa_fast_days	list	[]	—	ISO dates		Dates when scheduled food was withheld; tests whether anticipation persists without food delivery.
faa_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for faa; smaller values require stronger evidence.
sleep_immobility_minutes	float	1.0	—	minutes	≥ 0.1; ≤ 240	Continuous immobility duration required to score sleep; sampling must resolve this duration.
sleep_activity_threshold	float	0.0	—	recording units	≥ 0	Largest activity value still counted as immobile; zero suits exact-zero beam-break sensors.

Ultradian and temperature analyses

Key	Type	Default	Allowed values	Units	Constraints	What it changes
ultradian_min_hours	float	2.0	—	hours	≥ 0.05; ≤ 23; less_than:ultradian_max_hours	Shortest period in the faster-than-daily component search, separate from the circadian band.
ultradian_max_hours	float	8.0	—	hours	≥ 0.1; ≤ 23.5; greater_than:ultradian_min_hours	Longest period in the faster-than-daily component search.
ultradian_min_samples_per_cycle	float	8.0	—	dimensionless	≥ 2; ≤ 200	Samples per candidate cycle required before treating an ultradian peak as resolved.
ultradian_min_relative_amplitude	float	0.05	—	dimensionless	≥ 0; ≤ 1	Smallest ultradian amplitude relative to the circadian amplitude considered appreciable.
compensation_band_low	float	0.8	—	dimensionless	≥ 0.1; ≤ 1	Lower Q10 ratio classified as temperature compensated.
compensation_band_high	float	1.2	—	dimensionless	≥ 1; ≤ 10	Upper Q10 ratio classified as temperature compensated.
compensation_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for compensation; smaller values require stronger evidence.

Differential, paired, and predictive analyses

Key	Type	Default	Allowed values	Units	Constraints	What it changes
differential_compared	list	["mesor","amplitude","phase"]	mesor, amplitude, phase, period	dimensionless		Cosinor parameters permitted to differ between cohorts: mesor, amplitude, phase or period.
differential_free_period	bool	false	—	boolean switch		Add a between-cohort period term to the shared cosinor comparison model.
differential_rhythmic_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for differential rhythmic; smaller values require stronger evidence.
differential_min_units	int	3	—	count	≥ 2; ≤ 1000	Independent units per cohort required before comparing rhythm parameters.
differential_correction	str	"sidak"	bh, bonferroni, sidak, none	named choice or label		Multiple-testing correction within the differential comparison family, not across unrelated analyses.
differential_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for differential; smaller values require stronger evidence.
paired_min_subjects	int	3	—	count	≥ 2; ≤ 100000	Subjects present in both epochs required before a paired comparison.
paired_correction	str	"sidak"	bh, bonferroni, sidak, none	named choice or label		Multiple-testing correction within the paired comparison family, not across unrelated analyses.
paired_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for paired; smaller values require stronger evidence.
prediction_t_min_hours	float	20.0	—	hours	≥ 1; ≤ 240; less_than:prediction_t_max_hours	Shortest external cue period searched when predicting the entrainment range from a fitted response curve.
prediction_t_max_hours	float	28.0	—	hours	≥ 1; ≤ 240; greater_than:prediction_t_min_hours	Longest external cue period searched when predicting the entrainment range.
prediction_t_step_hours	float	0.1	—	hours	≥ 0.001; ≤ 6	Spacing between external cue periods in the entrainment prediction search.
prediction_max_cycles	int	60	—	cycles	≥ 1; ≤ 100000	Maximum simulated cycles before a transient is reported as not predicted to settle.
prediction_tolerance_hours	float	0.5	—	hours	≥ 1e-06; ≤ 12	Phase distance from the predicted steady state that counts as settled.
two_process_min_transitions	int	20	—	count	≥ 2; ≤ 100000	Observed sleep-wake transitions required before fitting the two-process sleep model.
two_process_min_days	int	5	—	days	≥ 1; ≤ 3650	Observed days required before fitting the two-process sleep model.
two_process_min_phases	int	4	—	count	≥ 1; ≤ 24	Occupied circadian phase bins required to constrain the sleep model.
two_process_max_iterations	int	2000	—	count	≥ 80; ≤ 1000000	Iteration budget for each two-process sleep-model optimiser restart.
two_process_restarts	int	4	—	count	≥ 1; ≤ 64	Independent starting searches used to expose ambiguous sleep-model fits.
two_process_step_minutes	float	10.0	—	minutes	≥ 1; ≤ 60	Time step used in the two-process sleep simulation.
two_process_parameter_span	float	1.5	—	dimensionless	≥ 1; ≤ 1000	Largest ratio across equally good parameter fits considered identifiable; wider spans are reported as ranges.
two_process_timing_tolerance	float	0.1	—	dimensionless	≥ 0; ≤ 1	Largest fraction of samples on which equally good model fits may disagree before predicted sleep timing is withheld.
two_process_mismatch_tie	float	0.02	—	dimensionless	≥ 0; ≤ 1	Difference in misclassified sample fractions below which sleep-model fits count as equally good.

Study, synchrony, spatial, coupling, dispersion, and power

Key	Type	Default	Allowed values	Units	Constraints	What it changes
study_max_recordings	int	200	—	count	≥ 1; ≤ 100000	Recording-count guard for a study run; increase explicitly for a genuinely larger study.
study_figure_formats	list	["svg"]	svg	dimensionless		Study figure formats; svg is supported and an empty list skips figures.
synchrony_min_cells	int	10	—	count	≥ 2; ≤ 100000	Cell traces required before reporting population phase concentration.
synchrony_min_contributing	int	5	—	count	≥ 2; ≤ 100000	Finite cell contributions required at each time point before reporting population concentration.
synchrony_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for synchrony; smaller values require stronger evidence.
spatial_min_points	int	6	—	count	≥ 4; ≤ 100000	Regions required before testing a spatial phase gradient rather than merely fitting a plane.
spatial_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for spatial; smaller values require stronger evidence.
spatial_permutations	int	2000	—	count	≥ 20; ≤ 100000	Phase-to-position shuffles used to test a spatial wave against scattered phases.
spatial_collinear_ratio	float	0.001	—	dimensionless	≥ 0; ≤ 1	Minimum smaller-to-larger position covariance eigenvalue ratio for identifying both spatial gradient directions.
coupling_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for coupling; smaller values require stronger evidence.
coupling_min_cycles	float	3	—	cycles	≥ 2; ≤ 100	Usable cycles required for a lag interval and lag-drift test between traces.
coupling_window_cycles	float	2	—	cycles	≥ 1; ≤ 100	Sliding-window width, in cycles, for the phase-locking series between traces.
period_dispersion_min_cycles	float	3	—	cycles	≥ 2; ≤ 100	Cycles required to separate period-driven coherence loss from initial phase spread.
period_dispersion_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for period dispersion; smaller values require stronger evidence.
power_simulations	int	400	—	count	≥ 20; ≤ 100000	Simulated experiments behind each point on the sample-size power curve.
power_max_n	int	64	—	count	≥ 3; ≤ 100000	Largest per-group sample size considered by the prospective power search.
power_target	float	0.8	—	dimensionless	≥ 0.01; ≤ 0.999	Power sought by the sample-size search, judged using the interval rather than the point estimate.
power_max_seconds	float	600.0	—	seconds	≥ 1; ≤ 86400	Predicted computing-time budget checked before a power simulation starts.
power_interval_alpha	float	0.05	—	dimensionless	≥ 1e-06; ≤ 0.5	Significance threshold for power interval; smaller values require stronger evidence.
power_seed	int	0	—	integer seed		Fixed random-generator seed for power; retain it to reproduce the stochastic calculation.

Spectral display, schedules, and colours

Key	Type	Default	Allowed values	Units	Constraints	What it changes
fft_blackman_harris	bool	false	—	boolean switch		Apply a Blackman-Harris taper before the Fourier spectrum to reduce spectral leakage.
fft_spectrum	str	"power"	amplitude, power	named choice or label		Display Fourier amplitude or squared-amplitude power; does not select a period estimator.
wavelet_cycles	float	6.0	—	cycles	≥ 3; ≤ 20	Cycles in the Morlet wavelet; trades temporal localisation against period resolution.
lights_on	float	7.0	—	clock hours		Clock hour when the repeating primary light interval starts; equal on and off times denote continuous darkness.
lights_off	float	19.0	—	clock hours		Clock hour when the repeating primary light interval ends.
lighting_schedule	list	[]	—	dated clock-hour intervals		Dated light-state intervals used instead of a repeating simple schedule.
zeitgeber_schedule	list	[]	—	dated cue intervals		Dated light, food, or other cue intervals used by entrainment analyses.
ld_start	str or null	null	—	ISO date or datetime		First date shown with the primary light-dark background on the actogram; display only.
ld_end	str or null	null	—	ISO date or datetime		Last date shown with the primary light-dark background on the actogram; display only.
dd_start	str or null	null	—	ISO date or datetime		First date shown with continuous darkness on the actogram; display only.
dd_end	str or null	null	—	ISO date or datetime		Last date shown with continuous darkness on the actogram; display only.
actogram_activity_color	str	"#111111"	—	hex colour		Colour of actogram activity marks; presentation only, independent of scientific estimates.
actogram_light_color	str	"#ffffff"	—	hex colour		Colour of actogram light marks; presentation only, independent of scientific estimates.
actogram_dark_color	str	"#d6d6d6"	—	hex colour		Colour of actogram dark marks; presentation only, independent of scientific estimates.
onset_color	str	"#c0392b"	—	hex colour		Colour of onset marks; presentation only, independent of scientific estimates.
offset_color	str	"#2878b5"	—	hex colour		Colour of offset marks; presentation only, independent of scientific estimates.
protocol_preset	str	"custom"	—	named choice or label		Name recorded in provenance for the selected settings preset.