Actogram¶

A raster-like chronobiological representation that aligns repeated cycles in successive rows so activity timing, entrainment, phase drift, and rhythmic disruption become visually comparable.

Version: v2 · 2026-09-06 · History
Domain-specific #: 1237
Origin domain: biology
Subdomain: chronobiology
Aliases: Activity actogram

Core Idea¶

An actogram is a chronobiological display that arranges observed activity or another rhythmic signal against time within a repeated cycle, with successive cycles placed on successive rows. Its purpose is to make phase, period, entrainment, drift, fragmentation, and transitions visible as geometric patterns. A mark in the plot stands for activity detected or summarized in a time bin; a blank region can stand for low activity, missing observations, or rest only after the encoding rule is declared. Jud and colleagues provide a detailed framework for analyzing rodent wheel-running records and show actograms as central representations for interpreting behavior under light–dark and constant conditions.^[1] The identity belongs to the cyclic layout and temporal alignment, not to wheel running itself.

Single-plotted actograms display each cycle once per row. Double-plotted actograms commonly display two consecutive cycles across each row and advance the start of the next row by one cycle. This duplicated context makes trajectories easier to follow across the row boundary, but each observation appears twice except at display edges; double plotting does not create additional data or double the sample size. Environmental schedules are often shown by light and dark bars, shading, or annotations. Under entrainment, an activity onset may remain vertically aligned at a stable phase relative to the schedule. Under constant conditions, a free-running period slightly different from the plotting cycle produces a diagonal drift. Pittendrigh and Daan's experimental series made such free-running and entrainment patterns central evidence in circadian theory.^[2]

The display is related to quantitative rhythm analysis but is not a period estimator by itself. Sokolove and Bushell's chi-square periodogram offers a statistical procedure for identifying candidate periodicities in biological time series and helps illustrate the division of labor: an actogram exposes phase organization and anomalies, while a periodogram summarizes periodic strength across tested periods.^[3] Visual slope can suggest a free-running period, yet bin size, plotting period, smoothing, missingness, masking by the environment, and observer choice can bias interpretation. The actogram should therefore be paired with explicit preprocessing, a declared cycle length, quantitative phase or period estimates where needed, and uncertainty.

The accepted catalog has Representation, Monitoring, Recurrence, Rhythm, Activity Diagram, and several time-series nodes. Activity Diagram is a process-modeling notation with a different identity. Monitoring concerns ongoing observation rather than the repeated-cycle visual encoding. The strict parent is Representation because the actogram maps observations and environmental schedule into a geometric display whose rows, columns, marks, and slope carry meaning. Its autonomous residual is the repeated-cycle alignment that converts phase evolution into spatial trajectory. That residual works for locomotor activity, sleep–wake state, physiological signals, or gene-expression rhythms without becoming a general graph type.

Structural Signature¶

Cyclic horizontal axis. Within-row position encodes time or phase inside a declared plotting cycle.
Successive rows. Each row advances by one cycle or another explicitly stated interval.
Activity encoding. Marks, height, density, or color summarize an observed variable in time bins.
Single or double plot convention. Duplication and row advance are declared so visual continuity is interpreted correctly.
Environmental annotation. Light–dark schedules, interventions, or condition changes can be aligned with observations.
Phase trajectory. Vertical alignment, diagonal drift, splitting, or fragmentation expresses timing behavior.
Plotting-period reference. Apparent slope is relative to the cycle used for display.
Preprocessing provenance. Binning, thresholding, smoothing, and missing-data handling define what a mark means.
Observation continuity. Gaps and excluded intervals are distinguished from genuine inactivity.
Interpretive pairing. Visual patterns can motivate quantitative phase, period, or entrainment analyses without replacing them.

What It Is Not¶

Not the raw sensor record. It is a representation produced from observations and preprocessing.
Not an activity diagram. That notation models workflows or processes, not biological phase across cycles.
Not a periodogram. It does not by itself test periodic power over candidate periods.
Not proof of an endogenous clock. Environmental masking and measurement artifacts can create apparent patterns.
Not necessarily wheel-running data. Many rhythmic variables can be encoded.
Not extra data when double plotted. Duplication aids continuity but does not add observations.
Not an automatic phase estimate. Onset and offset criteria still require definition and uncertainty.
Not evidence that blank means sleep. Absence of recorded activity has several interpretations.

Scope of Application¶

Actograms apply wherever repeated biological or behavioral cycles must be compared for phase structure. They are most established in circadian work but remain tied to declared observation and plotting conventions.

Locomotor rhythms. Wheel revolutions, infrared beam breaks, or movement summaries can be aligned across days.
Sleep–wake timing. State summaries can display consolidation, fragmentation, and phase shifts.
Hormonal or physiological rhythms. Repeated measurements can be rendered when sampling density supports the display.
Gene-expression rhythms. Time-aligned expression summaries can expose phase relationships across cycles.
Entrainment studies. Stable phase relative to an environmental cycle becomes geometrically visible.
Free-running conditions. Diagonal drift reveals mismatch between endogenous and plotting periods.
Phase-reset studies. Changes in trajectory before and after a perturbation can be compared descriptively.
Quality control. Missing intervals, sensor failures, and inconsistent preprocessing become visible when encoded honestly.

Clarity¶

State the measured variable, organism or system, clock convention, sampling interval, bin size, plotting cycle, row-advance convention, and whether the plot is single or double. Define exactly what mark height, darkness, or color means. Identify preprocessing such as thresholding, smoothing, normalization, or aggregation and distinguish zero from missing. Environmental schedules should include time reference and transitions rather than relying on unexplained black and white bars. If the plot uses zeitgeber time, local clock time, or circadian time, name the frame and any daylight-saving or time-zone treatment. A diagonal trajectory indicates phase change relative to the plotted cycle; deriving a period from slope requires the plotting convention and uncertainty. Double-plotted observations must not be counted twice in statistical analysis. Onset selection can be subjective in fragmented records, so report the rule or use an independent estimator. An actogram is strongest when the display, quantitative analysis, and original data remain traceably linked.

Manages Complexity¶

Long time series hide cyclic organization because adjacent samples emphasize short-term continuity rather than recurrence across days. The actogram folds time by a chosen cycle and stacks successive folds, converting phase evolution into a spatial path. Entrainment becomes near-vertical alignment; a period mismatch becomes diagonal drift; a phase shift becomes a break between trajectories; fragmentation becomes scattered or broadened marks. Environmental schedules can be aligned in the same frame. This representation separates several problems: sensors acquire observations, preprocessing defines activity summaries, layout exposes phase organization, and statistical methods estimate period or uncertainty. It also reveals ambiguity. A diagonal line can result from a genuine free-running rhythm or from a plotting-cycle mismatch; blank regions can mean rest or missing data; a sharp transition can be biological or a clock correction. By declaring the encoding and retaining source provenance, the display compresses complexity without turning visual salience into automatic inference.

Abstract Reasoning¶

Identify the observed variable, timestamp frame, sampling regularity, and missing-data pattern.
Choose a biologically and analytically justified plotting cycle rather than assuming exactly twenty-four hours.
Define binning, thresholding, normalization, and mark encoding before plotting.
Select single or double plotting and record how successive rows advance.
Align environmental schedules and condition changes to the same time frame.
Inspect vertical alignment, diagonal drift, fragmentation, and phase discontinuities as descriptive patterns.
Distinguish duplicated context in double plots from independent observations.
Test candidate period and phase claims with quantitative methods and uncertainty where conclusions require them.
Compare the display with raw data around gaps, transitions, and apparent anomalies.
Report the actogram as a representation whose interpretation depends on encoding and observation quality.

Knowledge Transfer¶

The transferable idea is to fold a time series by a reference period and align successive folds so phase becomes spatial geometry. Similar layouts can reveal periodic drift in machines, schedules, or communication traffic, but biological terms such as entrainment and free running require appropriate mechanisms. Representation is the strict parent because marks, rows, annotations, and slopes encode an underlying time series. The actogram's domain accent is repeated biological cycles, activity/rest signals, zeitgeber schedules, single/double plotting, and phase interpretation. Removing that accent leaves a periodic raster or folded time-series display rather than the exact node.

Examples¶

Canonical¶

A double-plotted display uses a twenty-four-hour reference cycle but shows forty-eight hours across each row. The next row begins twenty-four hours after the previous one, so most observations appear twice. Under constant darkness, activity onset shifts later each cycle and traces a rightward diagonal. The slope suggests an endogenous period longer than the plotting cycle, but a quantitative period estimate and confidence interval are calculated from the original time series rather than by treating duplicated marks as new data.

Mapped back: timestamped activity + declared cycle + double-row duplication → continuous phase trajectory → visually motivated, independently estimated free-running period.

Applied / In Practice¶

An entrainment study displays activity together with the light–dark schedule. Before a schedule change, onset is vertically aligned to lights-off. After the transition, the marks reorganize over several cycles and settle at a new phase. Missing sensor intervals are shaded separately from inactivity. The plot supports comparison of transient and steady phase, while a formal model quantifies the shift and excludes duplicated display data from the sample count.

Mapped back: activity summaries + environmental schedule + missingness encoding → phase-transition representation → scoped quantitative follow-up.

Structural Tensions¶

Visual continuity vs. duplicated evidence. Double plotting improves tracking by repeating data. Diagnostic: Is duplication excluded from inferential sample size?
Phase drift vs. plotting reference. Slope is relative rather than absolute. Diagnostic: Is the row cycle and time frame declared?
Inactivity vs. missingness. Both can appear blank. Diagnostic: Are absent observations encoded separately?
Rhythm vs. environmental masking. A pattern need not be endogenous. Diagnostic: Are condition and alternative drivers identified?
Display vs. estimator. A compelling diagonal is not a confidence interval. Diagnostic: Are numerical claims derived from independent analysis of original data?
Autonomous abstraction vs. Representation plus recurrence. Many plots show cycles. Diagnostic: Does the layout align successive biological cycles so phase evolution becomes a row-to-row trajectory?

Structural–Framed Character¶

Cyclic horizontal axis, successive-row advance, activity encoding, optional duplication, environmental annotation, phase trajectory, plotting reference, and preprocessing provenance are structural. Species, sensor, mark color, software, calendar date, and experimental condition are framed. A plot lacking repeated-cycle alignment may be a time-series display but not an actogram.

Structural Core vs. Domain Accent¶

The portable core is folding observations by a reference cycle and spatializing phase change. The domain accent is chronobiological activity, zeitgeber schedules, entrainment, free-running period, and actogram conventions. Removing the accent leaves Representation; retaining it yields the actogram identity.

Representation is the narrowest accepted prime because the actogram encodes observations, cycles, and environmental conditions into a geometric layout. Measurement supplies the data but does not own the repeated-row visual semantics; Recurrence is a depicted pattern, not the display architecture.

The prospective workspace queue contains one strict upward edge to prime:representation. No live DAG mutation is authorized.

Relationships to Other Abstractions¶

Current abstraction Actogram Domain-specific

Parents (1) — more general patterns this builds on

Actogram is a kind of Representation Prime

Representation is the narrowest accepted prime because the actogram encodes observations, cycles, and environmental conditions into a geometric layout.

Hierarchy path (1) — routes to 1 parentless root

Actogram → Representation → Abstraction

Neighborhood in Abstraction Space¶

Actogram sits in a sparse region of the domain-specific corpus (98^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (1565 abstractions)

Nearest neighbors

Continuous Cooling Transformation — 0.77
Reaching definition — 0.75
Flow process chart — 0.74
Log–log plot — 0.74
Horizon chart — 0.74

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With¶

Activity diagram. A workflow notation rather than a chronobiological display.
Periodogram. A statistical summary of power across candidate periods.
Raster plot. A broader event display that need not align repeated biological cycles.
Circadian rhythm. The phenomenon represented, not the plot.
Monitoring. Ongoing observation that may supply data without the actogram layout.
Sleep diary. A record format that can be visualized differently and carries self-report semantics.

References¶

[1] Claudia Jud, Iris Schmutz, Gabriela Hampp, Henrik Oster, and Urs Albrecht, ‘A Guideline for Analyzing Circadian Wheel-Running Behavior in Rodents under Different Lighting Conditions,’ Biological Procedures Online 7 (2005): 101–116, https://doi.org/10.1251/bpo109. registry ↩

[2] Colin S. Pittendrigh and Serge Daan, ‘A Functional Analysis of Circadian Pacemakers in Nocturnal Rodents,’ Journal of Comparative Physiology 106 (1976), series beginning https://doi.org/10.1007/BF01417856. registry ↩

[3] Philip G. Sokolove and William N. Bushell, ‘The Chi Square Periodogram: Its Utility for Analysis of Circadian Rhythms,’ Journal of Theoretical Biology 72, no. 1 (1978): 131–160, https://doi.org/10.1016/0022-5193(78)90022-X. registry ↩