Freedom251 › Physics Simulation Software
blobmodel
What blobmodel is
blobmodel is a free, open-source Python package for creating numerical realizations of moving and dissipating blobs in one or two spatial dimensions. It was developed by the Complex Systems Modelling group at UiT The Arctic University of Norway, originally for scrape-off-layer fusion research. Install it from PyPI with `pip install blobmodel` in an existing Python environment. The Model class accepts grid size, domain lengths, time step, total time, and initial time. Each realization is returned as an xarray dataset that can be saved to a named file. The package includes exponential, Lorentz, double exponential, Gaussian, secant, dipole, and rectangular blob shapes. DefaultBlobFactory supplies distributions for amplitudes, widths, velocities, pulse shapes, alignment, and tilt, while a custom BlobFactory can define parameters individually. Datasets can be displayed with `show_model`, including animation. Optional integer labels can mark occupied regions or individual blobs for supervised machine-learning datasets. One-dimensional mode removes perpendicular shape, and periodic y-boundaries wrap exiting blobs to the opposite edge.
Who it suits
It suits Python users modeling advecting and dissipating blobs in one or two dimensions, including fusion researchers and people preparing labeled machine-learning data.
What it does well
- Free and open-source Python package
- Supports one- and two-dimensional models
- Several built-in blob shapes
- Exports realizations as xarray datasets
- Optional labels support training datasets
What to know first
- Linux platform listed
- Custom factories require dimension and normalization care
- Periodic y-use requires a sufficiently large domain
Verdict
blobmodel provides configurable numerical blob simulations with multiple shapes, parameter distributions, visualization, and dataset output. Its documentation is particularly relevant to research and labeled data-generation workflows.
Facts from blobmodel's own pages
- Authoring tools
- Yes · blobmodel.readthedocs.io, 23 Sept 2026
- Blob shapes
- Supported shape choices include exponential, Lorentz, double exponential, Gaussian, secant, dipole, and rectangular. · blobmodel.readthedocs.io, 12 Oct 2026
- Custom parameters
- Users can implement a custom BlobFactory to define blob parameters individually; they are responsible for ensuring variables have correct dimensions and manually normalizing parameters if needed. · blobmodel.readthedocs.io, 11 Oct 2026
- Data export
- Yes · blobmodel.readthedocs.io, 23 Sept 2026
- Free plan
- Yes · blobmodel.readthedocs.io, 23 Sept 2026
- Installation
- The documentation provides installation from PyPI with `pip install blobmodel`. · blobmodel.readthedocs.io, 12 Oct 2026
- Intended research
- The project was originally developed to study the scrape-off layer of fusion experiments and may apply to other one- or two-dimensional systems. · github.com, 12 Oct 2026
- Label output
- Users can add same-region or per-blob integer labels to the output dataset, which the documentation notes may help create supervised machine-learning training data. · blobmodel.readthedocs.io, 12 Oct 2026
- Machine learning use
- The optional blob_labels field can label occupied regions or individual blobs, and the documentation says it can be useful for supervised machine learning training datasets. · blobmodel.readthedocs.io, 11 Oct 2026
- Model setup
- A model lets users specify grid dimensions, domain lengths, time step, time length, initial time, blob shape, drainage time, and blob count. · blobmodel.readthedocs.io, 12 Oct 2026
- One-dimensional mode
- Setting `one_dimensional` to true discards the perpendicular blob shape and sets `Ny=1` and `Ly=0`. · blobmodel.readthedocs.io, 12 Oct 2026
- Open source
- The documentation describes blobmodel as an open source project developed by the Complex Systems Modelling group at UiT The Arctic University of Norway. · blobmodel.readthedocs.io, 11 Oct 2026
- Output
- A realization is returned as an xarray dataset and can be saved locally by passing a filename to `make_realization`. · blobmodel.readthedocs.io, 12 Oct 2026
- Parameter distributions
- The default factory supports distributions for blob amplitudes, widths, velocities, pulse shapes, alignment, and tilt, and users can implement a custom factory. · blobmodel.readthedocs.io, 12 Oct 2026
- Periodic boundary behavior
- With `periodic_y` enabled, blobs leaving the domain in the y direction enter at the opposite end; the docs caution that the y domain should be large compared with blob widths. · blobmodel.readthedocs.io, 11 Oct 2026
- Periodic boundary caveat
- The documentation advises using y-periodicity only when the y-domain is large compared with blob widths and says the code warns when a blob width is less than `0.1 * Ly`. · blobmodel.readthedocs.io, 12 Oct 2026
- Purpose
- blobmodel is a Python package for creating numerical realizations of advecting and dissipating blobs in up to two spatial dimensions. · blobmodel.readthedocs.io, 12 Oct 2026
- Support and contribution
- The project invites users to report bugs or request features on GitHub and contribute pull requests. · blobmodel.readthedocs.io, 12 Oct 2026
- 3D visuals
- No · blobmodel.readthedocs.io, 23 Sept 2026
- Visualization
- The package provides `show_model` to plot a model realization as an animation. · blobmodel.readthedocs.io, 12 Oct 2026
Where blobmodel stands on Freedom251
- Physics Simulation Software#65 of 89