v4.1.0

VERSION 4.1.0, Jhangri

Jhangri is an Indian sweet in the shape of a flower. It is made of white-lentil (Vigna mungo) batter, deep-fried in ornamental shape to form the crunchy, golden body, which is then soaked in sugar syrup lightly flavoured with spices.

This release has the following major changes:

• Improved support for reading NeuroML2 models
• HHGate2D: separate xminA, xminB, etc. for A and B tables replaced by single xmin, xmax, xdivs, ymin, ymax, and ydivs fields for both tables.
• Build system switched from cmake to meson
• Native binaries for Windows
• Updated C/C++ code to conform to c/c++-17 standard
• Various bugfixes

Dependencies

Although moose depends on gsl>=1.16, the binaries are linked against a more recent GSL, and require gsl-2.7 or gsl-2.8 to be installed on the system. The easiest way to set this up is to use conda/mamba/micromamba, with the channel conda-forge.

1. Install micromamba
2. Create an environment with the dependencies: micromamba create -n moose gsl hdf5 numpy matplotlib vpython -c conda-forge
3. Activate the moose environment: micromamba activate moose
4. Download the file (under Assets) matching the python version of this environment (e.g., cp311 for CPython 3.11), OS, and CPU architecture. For Linux/WSL, use pymoose-*-manylinux_2_28_*.whl
5. Install with pip while in the same directory as the downloaded file: pip install <filename>.

See README.md and INSTALL.md for more detailed instructions on installation and requirements.

Check installation

Test it with

python -c "import moose; ch = moose.HHChannel('ch'); moose.le()"

and it should print something like:

Elements under /
    /Msgs
    /clock
    /classes
    /postmaster
    /ch

Moose_4.0.0 or Jalebi

Jalebi is an Indian sweet involving a golden twisting tube like a hyper-pretzel, of crunchy batter soaked in sugar syrup lightly flavoured with spices and sometimes lemon.

This release has the following major changes:

A major under-the-hood change to numerics for chemical calculations, eliminating the use of 'zombie' objects for the solvers. This simplifies and cleans up the code and object access, but doesn't alter runtimes.

Another major under-the-hood change to use pybind11 as a much cleaner way to interface the parser with the C++ numerical code.

Addition of a thread-safe and faster parser based on ExprTK

Resurrected objects for handling simulation output saving using HDF5 format. There is an HDFWriter class, an NSDFWriter, and a new NSDFWriter2. The latter two implement storage in NSDF, Neuronal Simulation Data Format, Ray et al Neuroinformatics 2016. NSDF is built on HDF5 and builds up a specification designed to ensure ready replicability as well as self- description of model output.

Multiple enhancements to rdesigneur, including vastly improved 3-D graphics output using VPython.

Various bugfixes

maintenance release_v3.1.5

Merge pull request #382 from dilawar/chamcham

Maintenance release 3.1.5

3.1.4 (Chamcham Series)

• pypi wheels for python3.
• Minor tweaks in build system.
• Bug fixes related to Python2/Python3 compatibility.

Chamcham (3.1.3): Bug fix release

• Python wheels.
• Fixes to build system.
• Bugfixes.

Chamcham (3.1.2) - bugfix and enhancement over 3.1.1

• Improvement to GUI
• Improvement to redesigneur
• Bug-fixes.
• Boost ODE solver are beta now.

Chamcham (3.1.1) - bugfix and enhancement over 3.1.0

• SBML support is at python level.
• Enhanced rdesigneur interface.
• Experimental support for parallel execution of biochemical reaction-diffusion system.
• Various bug fixes in github repo for BhallaLab/moose-core and BhallaLab/moose-gui

Pre release 3.1.0

• Streamer support. Store large simulation data in csv and npy (beta) format.
• Enhanced rdesigneur
• bugfixes
• Faster solvers using boost libraries.

Ghevar beta.1 release.

Get the downloadable files for https://github.com/BhallaLab/moose/wiki/Install