FPGA Hardware - Mycelium Growth Dataset

(Readme) Zenodo Dataset: FPGA Hardware - Mycelium Growth DatasetDataset root folder: Zenodo/ 1. OverviewThis dataset contains outcomes of mycelium growth patterns produced by repeatedly running a reaction–diffusion (RD) model on a two-dimensional spatial domain. Each run evolves a coupled set of fields under controlled and variable environmental conditions and stores the final state after 50,000 iterations. Herein, four spatially distinct mycelium instances are produced, corresponding to four environmental scenarios and four distinct mycelium types. The spatial domain is partitioned into multiple equal subareas, with four selected subareas actively seeded and analyzed. Each seeded subarea is associated with a specific environmental factor and a distinct set of reaction–diffusion parameters. All seeded subareas evolve simultaneously within the same run while preserving identical initialization conditions. This enables parallel mycelium growth in the FPGA hardware. Afterwards, the final biomass field is spatially partitioned, and the four subarea-specific mycelium instances are reorganized into four independent dataset branches (mycelium_1 to mycelium_4). Each branch represents a distinct mycelium type driven primarily by a single environmental factor (temperature, humidity, or light consisting of light wavelength and intensity). The final-state images are stored in this dataset, exported in JPEG format. 2. Folder StructureZenodo/ mycelium_1/ FPGA Sample T<TemperatureValue>/ c_50000.jpeg FPGA Sample T<TemperatureValue>/ c_50000.jpeg ... mycelium_2/ FPGA Sample H<HumidityValue>/ c_50000.jpeg FPGA Sample H<HumidityValue>/ c_50000.jpeg ... mycelium_3/ FPGA Sample L1_I<Intensity1Value>_W<Wavelength1Value>/ c_50000.jpeg ... mycelium_4/ FPGA Sample L2_I<Intensity2Value>_W<Wavelength2Value>/ c_50000.jpeg ... 3. Model VariablesThe reaction–diffusion system evolves three coupled state variables on the spatial grid:- u(x,y,t): an activator field associated with growth-driving processes- v(x,y,t): an suppressor field regulating suppression- c(x,y,t): a bistable biomass field belonging in the interval [0,1] The dataset visualizes the final state of the biomass field c, which represents the spatial morphology of the developed mycelial network within each subarea. 4. Environmental Parameter Sampling and Subarea AssignmentFor each run, environmental parameters are sampled independently from predefined continuous ranges:- Temperature: 10–35- Humidity: 70–100- Light intensity: 0–100- Wavelength: 350–700 Two independent intensity–wavelength pairs are generated per run to define two distinct light-sensitive subareas. A base substrate matrix n is initialized as a spatially heterogeneous field with randomly sampled values within the unit interval, perturbed by additive Gaussian noise and clipped to remain within these bounds. The sampled environmental parameters are transformed into nonlinear environmental scaling factors. Each scaling factor is applied multiplicatively to the corresponding subarea of the substrate matrix, yielding a single aggregated substrate field n that encodes all environmental influences in a spatially resolved manner. This approach ensures that all subareas share identical initialization conditions while being subject to distinct environmental modulation. Each run is initialized with identical localized seed conditions in all active subareas and evolved for a fixed number of time steps. Repeating this procedure across many runs produces a structured dataset that combines systematic single-factor isolation, distinct mycelium types and their parallel growth, and randomized multi-factor environmental variability.