1. Dataset Description

This dataset contains output from the BYTE (B3GET Yielded Thesis Experiment) simulations, generated using the B3GET (Behavior, Ecology, Genetics, Evolution, and Tradeoffs) agent-based model, version 1.2.0. B3GET is implemented in NetLogo, a multi-agent programmable modeling environment widely used for simulating complex systems. B3GET simulates ANIMAL agents living in two-dimensional PLANT landscapes, enabling studies of behavioral ecology and evolution in spatially explicit, energetically constrained environments. BYTE consists of 4,608 simulations across four virtual primate species (chimpanzees, geladas, hamadryas baboons, and olive baboons) under diverse environmental and model conditions.

Each simulation produced multi-level outputs capturing gene-level, individual-level, group-level, population-level, and ecosystem-level dynamics. Recorded data includes state variable values at global (e.g., plant-seasonality) and individual (e.g., body.size) levels, as well as numerous measurements of emergent simulation properties (e.g., current energy availability, mean body.size of adult males). The dataset supports a wide range of analyses, including survival, mating, reproduction, sociality, aggression, dispersal, kin selection, growth and development, foraging, and ecological structure.

2. File Descriptions

The simulation output data are grouped by file type and compressed into the following ZIP archives:

• BYTE_outputs_focal_actions_[world].zip: Detailed timestep-by-timestep interactions and behaviors of randomly selected focal subjects.

• BYTE_outputs_focal_scans.zip: Timestep-by-timestep states of focal subjects.

• BYTE_outputs_genotypes.zip: Behavioral gene values across populations over time.

• BYTE_outputs_group_scans.zip: Group-level metrics (size, composition, turnover) over time.

• BYTE_outputs_individuals_[population].zip: Individual-level summaries recorded at the end of each individual's life.

• BYTE_outputs_netlogo_saves_[world].zip: NetLogo files to reload simulation endpoints for further analysis or continuation.

• BYTE_outputs_runtime_verification.zip: Runtime validation test results.

• BYTE_outputs_simulation_scans.zip: Simulation-level ecological and demographic metrics over time.

• BYTE_outputs_summary.zip: High-level summaries and patterns across simulations.

• BYTE_outputs_views.zip / BYTE_outputs_territories.zip: Images showing spatial dynamics (e.g., habitat structure, group home ranges).

Additional metadata and documentation files:

• b3get_v1.2.0.zip: Archived source code for the B3GET model, version 1.2.0, used to generate all BYTE simulations.

• BYTE_inputs.zip: Population and genotype data used to initialize BYTE simulations.

• manifest.csv: Provides file sizes, counts, and description of each ZIP archive and its contents.

• data_dictionary.csv: Defines all variables/columns in the simulation output data.

• README.txt: This file.

3. Simulation Naming Scheme

Simulations are grouped into worlds labeled A–P, each defining a unique combination of global parameter settings. Within each world, simulation settings differ only by population (chimpanzees, geladas, hamadryas, olives) and plant neighbor parameters (plant-minimum-neighbors, plant-maximum-neighbors). ZIP archives contain data files organized by simulation ID, simulation world, and/or population name.

Simulation IDs follow the format: sWPLHR, where:

• W: Simulation world (A–P)
• P: Population (C = chimpanzees, G = geladas, H = hamadryas, O = olives)
• L: Value of plant-minimum-neighbors (0–9)
• H: Value of plant-maximum-neighbors (0–9)
• R: Run replicate (A or B)

4. Explanation of Units

Energy Units: A custom unit of measurement in the B3GET model representing energy reserves or resource availability for entities (ANIMALs and PLANTs). This is a relative, abstract scale used internally for energy accounting and ecological interactions.

Ticks: Fundamental time unit in NetLogo representing discrete timesteps. Each tick advances the simulation forward by one unit of time, during which agent states and environmental variables may be updated in response to actions taken by the agents.

Cells: The spatial unit within NetLogo’s grid-based environment. Each cell contains exactly one PLANT entity, and multiple ANIMAL entities may occupy and move between these cells as they locomote. Cell dimensions are fixed by simulation settings and remain consistent across the BYTE simulations.

5. Missing Values, Empty Cells, and List Structures

The core simulation input and output CSV files are preserved exactly as generated to ensure reproducibility. Some files intentionally contain empty cells or list-formatted entries. Empty cells indicate values that are structurally not applicable to a given record; these values are not data errors and should not be interpreted as missing due to data loss. To reduce ambiguity for readers, documentation files (README, data_dictionary.csv, manifest.csv) explicitly describe when values or data are not applicable, not generated, or structurally empty.

Empty square brackets ([]), where present, denote valid empty lists. These occur in list-type variables when no items were recorded (e.g., no interaction partners, no carried items). An empty list is a meaningful value distinct from “missing” data.

6. Data Dictionary Details

Variables in the data dictionary are categorized to clarify their roles within the simulation:

• global static: Fixed global parameters or settings unchanged during a simulation.

• global dynamic: Global conditions or measurements that change throughout the simulation.

• ANIMAL static: Individual-level traits fixed at conception and unchanged throughout an individual's lifetime.

• ANIMAL dynamic: Individual-level traits, statuses, or measurements that can change as the individual develops and ages.

• ANIMAL action: Behaviors or activities performed by or directed toward individuals, including themselves, which can occur multiple times per timestep.

• ANIMAL group: Properties related to the social groups in the simulation, including affiliation, group size, and relatedness measures among individuals.

Brackets [ ]: Brackets around text in a variable name indicate a placeholder for multiple closely related variables that differ only by a specified variant. For example, plant-abundance-[statistic] includes plant-abundance-mean, plant-abundance-minimum as options.

Forward Slash /: A forward slash between terms in a variable name (e.g., base-litter-size / maximum-litter-size) indicates multiple acceptable naming options for the same conceptual variable.

7. License

This dataset is released under the Public Domain (CC0 1.0) license waiver. You are free to use, modify, and distribute the data without restriction.

8. Citation

DOI: 10.5061/dryad.d2547d8d4