Purpose

Wrap the righ-hand-side of the bunch1s models one by one with f2py.

Usage

Download the latest build library built for your platform (the code is automatically compiled when pushed to github):

Windows

Linux

Then include it your python script with:

import bunch1s

Currently provides the following equation right hand sides:

python3 -c "import bunch1s; print(bunch1s.__doc__);"
...
This module 'bunch1s' is auto-generated with f2py (version:1.24.0).
Functions:
    dy = gpmm2(y,bdef,p1,p2,m=shape(y, 0))
    dy = g1smm(y,bdef,p1,p2,m=shape(y, 0))
    dy = g1slw(y,par,m=shape(y, 0))
    dy = gkrug(y,par,m=shape(y, 0))
    dy = g_pk2(y,par,m=shape(y, 0))
    dy = g_mm0(y,par,m=shape(y, 0))
    dy = g_mm1(y,par,m=shape(y, 0))
    dy = gise2(y,par,m=shape(y, 0))

Concept notes

We re-format the right-hand-sides to be compatible with the .f90 (Fortan-90) standart as it's more flexible with formatting and automated formatters and linters (fprettify) exist that make the source code more readable and maintainable.

Compiling manually

To compile manually clone this repository and install: cmake, python, gfortran, g++, gcc, make (provided by mingw in windows).

Open a terminal in the root of the repository and run:

$ mkdir build
$ cmake -S . -B build/ -G "Unix Makefiles"
$ cd build
$ make

CMake will autoconfigure all libraries, dependencies and compilers and emit valid unix makefiles which upon invoking make inside the build folder will compile the project successfully.

Initial results

Integrating the wrapped g1smm with LSODA from SciPy for parameters:

bdef = 2.0
p1 = 3.0
p2 = 2.0

Yields nice looking trajectories and grouping:

pybunch1s and integrate_trajectories

pybunch1s provides a high-level python interface for all the wrapped functions in order to enable type-checking and IDE suggestions.

integrate_trajectories is now a library that provides all the basic tools needed to integrate a model.

surface_statistics

Provides lstat and hstat