pyodesys
========
.. image:: http://hackspett.bjodah.se/api/badges/8/status.svg
:target: http://hackspett.bjodah.se/repos/8
:alt: Build status on Woodpecker CI
.. image:: https://img.shields.io/pypi/v/pyodesys.svg
:target: https://pypi.python.org/pypi/pyodesys
:alt: PyPI version
.. image:: https://img.shields.io/pypi/l/pyodesys.svg
:target: https://github.com/bjodah/pyodesys/blob/master/LICENSE
:alt: License
.. image:: http://artifacts.bjodah.se/pyodesys/branches/master/htmlcov/coverage.svg
:target: http://artifacts.bjodah.se/pyodesys/branches/master/htmlcov
:alt: coverage
.. image:: http://joss.theoj.org/papers/10.21105/joss.00490/status.svg
:target: https://doi.org/10.21105/joss.00490
:alt: Journal of Open Source Software DOI
``pyodesys`` provides a straightforward way
of numerically integrating systems of ordinary differential equations (initial value problems).
It unifies the interface of several libraries for performing the numerical integration as well as
several libraries for symbolic representation. It also provides a convenience class for
representing and integrating ODE systems defined by symbolic expressions, e.g. `SymPy <http://www.sympy.org>`_
expressions. This allows the user to write concise code and rely on ``pyodesys`` to handle the subtle differences
between libraries.
The numerical integration is performed using either:
- `scipy.integrate.ode <http://docs.scipy.org/doc/scipy/reference/generated/scipy.integrate.ode.html>`_
- pygslodeiv2_
- pyodeint_
- pycvodes_
.. _pygslodeiv2: https://github.com/bjodah/pygslodeiv2
.. _pyodeint: https://github.com/bjodah/pyodeint
.. _pycvodes: https://github.com/bjodah/pycvodes
Note that implicit steppers require a user supplied callback for calculating the Jacobian.
``pyodesys.SymbolicSys`` derives the Jacobian automatically.
The symbolic representation is usually in the form of SymPy expressions, but the user may
choose another symbolic back-end (see `sym <https://github.com/bjodah/sym>`_).
When performance is of utmost importance, e.g. in model fitting where results are needed
for a large set of initial conditions and parameters, the user may transparently
rely on compiled native code (classes in ``pyodesys.native.native_sys`` can generate optimal C++ code).
The major benefit is that there is no need to manually rewrite the corresponding expressions in another
programming language.
Documentation
-------------
Auto-generated API documentation for latest stable release is found here:
`<https://bjodah.github.io/pyodesys/latest>`_
(and the development version for the current master branch is found here:
`<http://artifacts.bjodah.se/pyodesys/branches/master/html>`_).
Installation
------------
Simplest way to install pyodesys and its (optional) dependencies is to use the
`conda package manager <http://conda.pydata.org/docs/>`_:
::
$ conda install -c bjodah pyodesys pytest
$ python -m pytest --pyargs pyodesys
Optional dependencies
~~~~~~~~~~~~~~~~~~~~~
If you used ``conda`` to install pyodesys_ you can skip this section.
But if you use ``pip`` you may want to know that the default installation
of ``pyodesys`` only requires SciPy::
$ pip install pyodesys
$ pytest --pyargs pyodesys -rs
The above command should finish without errors but with some skipped tests.
The reason for why some tests are skipped should be because missing optional solvers.
To install the optional solvers you will first need to install third party libraries for
the solvers and then their python bindings. The 3rd party requirements are as follows:
- pygslodeiv2_ (requires GSL_ >=1.16)
- pyodeint_ (requires boost_ >=1.72.0)
- pycvodes_ (requires SUNDIALS_ >=5.3.0)
.. _GSL: https://www.gnu.org/software/gsl/
.. _boost: http://www.boost.org/
.. _SUNDIALS: https://computation.llnl.gov/projects/sundials
if you want to see what packages need to be installed on a Debian based system you may look at this
`Dockerfile <scripts/environment/Dockerfile>`_.
If you manage to install all three external libraries you may install pyodesys with the option "all"::
$ pip install pyodesys[all]
$ pytest --pyargs pyodesys -rs
now there should be no skipped tests. If you try to install pyodesys on a machine where you do not have
root permissions you may find the flag ``--user`` helpful when using pip. Also if there are multiple
versions of python installed you may want to invoke python for an explicit version of python, e.g.::
$ python3.6 -m pip install --user pyodesys[all]
see `setup.py <setup.py>`_ for the exact list of requirements.
Using Docker
~~~~~~~~~~~~
If you have `Docker <https://www.docker.com>`_ installed, you may use it to host a jupyter
notebook server::
$ ./scripts/host-jupyter-using-docker.sh . 8888
the first time you run the command some dependencies will be downloaded. When the installation
is complete there will be a link visible which you can open in your browser. You can also run
the test suite using the same docker-image::
$ ./scripts/host-jupyter-using-docker.sh . 0
there will be one skipped test (due to symengine missing in this pip installed environment) and
quite a few instances of RuntimeWarning.
Examples
--------
The classic van der Pol oscillator (see `examples/van_der_pol.py <examples/van_der_pol.py>`_)
.. code:: python
>>> from pyodesys.symbolic import SymbolicSys
>>> def f(t, y, p):
... return [y[1], -y[0] + p[0]*y[1]*(1 - y[0]**2)]
...
>>> odesys = SymbolicSys.from_callback(f, 2, 1)
>>> xout, yout, info = odesys.integrate(10, [1, 0], [1], integrator='odeint', nsteps=1000)
>>> _ = odesys.plot_result()
>>> import matplotlib.pyplot as plt; plt.show() # doctest: +SKIP
.. image:: https://raw.githubusercontent.com/bjodah/pyodesys/master/examples/van_der_pol.png
If the expression contains transcendental functions you will need to provide a ``backend`` keyword argument:
.. code:: python
>>> import math
>>> def f(x, y, p, backend=math):
... return [backend.exp(-p[0]*y[0])] # analytic: y(x) := ln(kx + kc)/k
...
>>> odesys = SymbolicSys.from_callback(f, 1, 1)
>>> y0, k = -1, 3
>>> xout, yout, info = odesys.integrate(5, [y0], [k], integrator='cvode', method='bdf')
>>> _ = odesys.plot_result()
>>> import matplotlib.pyplot as plt
>>> import numpy as np
>>> c = 1./k*math.exp(k*y0) # integration constant
>>> _ = plt.plot(xout, np.log(k*(xout+c))/k, '--', linewidth=2, alpha=.5, label='analytic')
>>> _ = plt.legend(loc='best'); plt.show() # doctest: +SKIP
.. image:: https://raw.githubusercontent.com/bjodah/pyodesys/master/examples/lnx.png
If you already have symbolic expressions created using e.g. SymPy you can create your system from those:
.. code:: python
>>> import sympy as sp
>>> t, u, v, k = sp.symbols('t u v k')
>>> dudt = v
>>> dvdt = -k*u # differential equations for a harmonic oscillator
>>> odesys = SymbolicSys([(u, dudt), (v, dvdt)], t, [k])
>>> result = odesys.integrate(7, {u: 2, v: 0}, {k: 3}, integrator='gsl', method='rk8pd', atol=1e-11, rtol=1e-12)
>>> _ = plt.subplot(1, 2, 1)
>>> _ = result.plot()
>>> _ = plt.subplot(1, 2, 2)
>>> _ = plt.plot(result.xout, 2*np.cos(result.xout*3**0.5) - result.yout[:, 0])
>>> plt.show() # doctest: +SKIP
.. image:: https://raw.githubusercontent.com/bjodah/pyodesys/master/examples/harmonic.png
You can also refer to the dependent variables by name instead of index:
.. code:: python
>>> odesys = SymbolicSys.from_callback(
... lambda t, y, p: {
... 'x': -p['a']*y['x'],
... 'y': -p['b']*y['y'] + p['a']*y['x'],
... 'z': p['b']*y['y']
... }, names='xyz', param_names='ab', dep_by_name=True, par_by_name=True)
...
>>> t, ic, pars = [42, 43, 44], {'x': 7, 'y': 5, 'z': 3}, {'a': [11, 17, 19], 'b': 13}
>>> for r, a in zip(odesys.integrate(t, ic, pars, integrator='cvode'), pars['a']):
... assert np.allclose(r.named_dep('x'), 7*np.exp(-a*(r.xout - r.xout[0])))
... print('%.2f ms ' % (r.info['time_cpu']*1e3)) # doctest: +SKIP
...
10.54 ms
11.55 ms
11.06 ms
Note how we generated a list of results for each value of the parameter ``a``. When using a class
from ``pyodesys.native.native_sys`` those integrations are run in separate threads (bag of tasks
parallelism):
.. code:: python
>>> from pyodesys.native import native_sys
>>> native = native_sys['cvode'].from_other(odesys)
>>> for r, a in zip(native.integrate(t, ic, pars), pars['a']):
... assert np.allclose(r.named_dep('x'), 7*np.exp(-a*(r.xout - r.xout[0])))
... print('%.2f ms ' % (r.info['time_cpu']*1e3)) # doctest: +SKIP
...
0.42 ms
0.43 ms
0.42 ms
For this small example we see a 20x (serial) speedup by using native code. Bigger systems often see 100x speedup.
Since the latter is run in parallel the (wall clock) time spent waiting for the results is in practice
further reduced by a factor equal to the number of cores of your CPU (number of threads used is set by
the environment variable ``ANYODE_NUM_THREADS``).
For further examples, see `examples/ <https://github.com/bjodah/pyodesys/tree/master/examples>`_, and rendered
jupyter notebooks here: `<http://artifacts.bjodah.se/pyodesys/branches/master/examples>`_
Run notebooks using binder
~~~~~~~~~~~~~~~~~~~~~~~~~~
Using only a web-browser (and an internet connection) it is possible to explore the
notebooks here: (by the courtesy of the people behind mybinder)
.. image:: http://mybinder.org/badge.svg
:target: https://mybinder.org/v2/gh/bjodah/pyodesys/v0.11.6?filepath=index.ipynb
:alt: Binder
Citing
------
If you make use of pyodesys in e.g. academic work you may cite the following peer-reviewed publication:
.. image:: http://joss.theoj.org/papers/10.21105/joss.00490/status.svg
:target: https://doi.org/10.21105/joss.00490
:alt: Journal of Open Source Software DOI
Depending on what underlying solver you are using you should also cite the appropriate paper
(you can look at the list of references in the JOSS article). If you need to reference,
in addition to the paper, a specific point version of pyodesys (for e.g. reproducibility)
you can get per-version DOIs from the zenodo archive:
.. image:: https://zenodo.org/badge/43131469.svg
:target: https://zenodo.org/badge/latestdoi/43131469
:alt: Zenodo DOI
Licenseing
----------
The source code is Open Source and is released under the simplified 2-clause BSD license. See `LICENSE <LICENSE>`_ for further details.
Contributing
------------
Contributors are welcome to suggest improvements at https://github.com/bjodah/pyodesys (see further details `here <CONTRIBUTORS.rst>`_).
Author
------
Original author: Björn I. Dahlgren (gmail address: bjodah).
See file `AUTHORS <AUTHORS>`_ for a list of all authors.
Raw data
{
"_id": null,
"home_page": "https://github.com/bjodah/pyodesys",
"name": "pyodesys",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.7",
"maintainer_email": null,
"keywords": null,
"author": "Bjoern I. Dahlgren",
"author_email": "bjodah@gmail.com",
"download_url": "https://files.pythonhosted.org/packages/ca/dc/15454bd8d120e10105b36a9321b0dd9ee15e2b1533f61d53261cd14399a6/pyodesys-0.14.3.tar.gz",
"platform": null,
"description": "pyodesys\n========\n\n.. image:: http://hackspett.bjodah.se/api/badges/8/status.svg\n :target: http://hackspett.bjodah.se/repos/8\n :alt: Build status on Woodpecker CI\n.. image:: https://img.shields.io/pypi/v/pyodesys.svg\n :target: https://pypi.python.org/pypi/pyodesys\n :alt: PyPI version\n.. image:: https://img.shields.io/pypi/l/pyodesys.svg\n :target: https://github.com/bjodah/pyodesys/blob/master/LICENSE\n :alt: License\n.. image:: http://artifacts.bjodah.se/pyodesys/branches/master/htmlcov/coverage.svg\n :target: http://artifacts.bjodah.se/pyodesys/branches/master/htmlcov\n :alt: coverage\n.. image:: http://joss.theoj.org/papers/10.21105/joss.00490/status.svg\n :target: https://doi.org/10.21105/joss.00490\n :alt: Journal of Open Source Software DOI\n\n``pyodesys`` provides a straightforward way\nof numerically integrating systems of ordinary differential equations (initial value problems).\nIt unifies the interface of several libraries for performing the numerical integration as well as\nseveral libraries for symbolic representation. It also provides a convenience class for \nrepresenting and integrating ODE systems defined by symbolic expressions, e.g. `SymPy <http://www.sympy.org>`_\nexpressions. This allows the user to write concise code and rely on ``pyodesys`` to handle the subtle differences\nbetween libraries.\n\nThe numerical integration is performed using either:\n\n- `scipy.integrate.ode <http://docs.scipy.org/doc/scipy/reference/generated/scipy.integrate.ode.html>`_\n- pygslodeiv2_\n- pyodeint_\n- pycvodes_\n\n.. _pygslodeiv2: https://github.com/bjodah/pygslodeiv2 \n.. _pyodeint: https://github.com/bjodah/pyodeint\n.. _pycvodes: https://github.com/bjodah/pycvodes\n\n\nNote that implicit steppers require a user supplied callback for calculating the Jacobian.\n``pyodesys.SymbolicSys`` derives the Jacobian automatically.\n\nThe symbolic representation is usually in the form of SymPy expressions, but the user may\nchoose another symbolic back-end (see `sym <https://github.com/bjodah/sym>`_).\n\nWhen performance is of utmost importance, e.g. in model fitting where results are needed\nfor a large set of initial conditions and parameters, the user may transparently\nrely on compiled native code (classes in ``pyodesys.native.native_sys`` can generate optimal C++ code).\nThe major benefit is that there is no need to manually rewrite the corresponding expressions in another\nprogramming language.\n\nDocumentation\n-------------\nAuto-generated API documentation for latest stable release is found here:\n`<https://bjodah.github.io/pyodesys/latest>`_\n(and the development version for the current master branch is found here:\n`<http://artifacts.bjodah.se/pyodesys/branches/master/html>`_).\n\n\nInstallation\n------------\nSimplest way to install pyodesys and its (optional) dependencies is to use the\n`conda package manager <http://conda.pydata.org/docs/>`_:\n\n::\n\n $ conda install -c bjodah pyodesys pytest\n $ python -m pytest --pyargs pyodesys\n\nOptional dependencies\n~~~~~~~~~~~~~~~~~~~~~\nIf you used ``conda`` to install pyodesys_ you can skip this section.\nBut if you use ``pip`` you may want to know that the default installation\nof ``pyodesys`` only requires SciPy::\n\n $ pip install pyodesys\n $ pytest --pyargs pyodesys -rs\n\nThe above command should finish without errors but with some skipped tests.\nThe reason for why some tests are skipped should be because missing optional solvers.\nTo install the optional solvers you will first need to install third party libraries for\nthe solvers and then their python bindings. The 3rd party requirements are as follows:\n\n- pygslodeiv2_ (requires GSL_ >=1.16)\n- pyodeint_ (requires boost_ >=1.72.0)\n- pycvodes_ (requires SUNDIALS_ >=5.3.0)\n\n.. _GSL: https://www.gnu.org/software/gsl/\n.. _boost: http://www.boost.org/\n.. _SUNDIALS: https://computation.llnl.gov/projects/sundials\n\nif you want to see what packages need to be installed on a Debian based system you may look at this\n`Dockerfile <scripts/environment/Dockerfile>`_.\n\nIf you manage to install all three external libraries you may install pyodesys with the option \"all\"::\n\n $ pip install pyodesys[all]\n $ pytest --pyargs pyodesys -rs\n\nnow there should be no skipped tests. If you try to install pyodesys on a machine where you do not have\nroot permissions you may find the flag ``--user`` helpful when using pip. Also if there are multiple\nversions of python installed you may want to invoke python for an explicit version of python, e.g.::\n\n $ python3.6 -m pip install --user pyodesys[all]\n\nsee `setup.py <setup.py>`_ for the exact list of requirements.\n\nUsing Docker\n~~~~~~~~~~~~\nIf you have `Docker <https://www.docker.com>`_ installed, you may use it to host a jupyter\nnotebook server::\n\n $ ./scripts/host-jupyter-using-docker.sh . 8888\n\nthe first time you run the command some dependencies will be downloaded. When the installation\nis complete there will be a link visible which you can open in your browser. You can also run\nthe test suite using the same docker-image::\n\n $ ./scripts/host-jupyter-using-docker.sh . 0\n\nthere will be one skipped test (due to symengine missing in this pip installed environment) and\nquite a few instances of RuntimeWarning.\n\nExamples\n--------\nThe classic van der Pol oscillator (see `examples/van_der_pol.py <examples/van_der_pol.py>`_)\n\n.. code:: python\n\n >>> from pyodesys.symbolic import SymbolicSys\n >>> def f(t, y, p):\n ... return [y[1], -y[0] + p[0]*y[1]*(1 - y[0]**2)]\n ... \n >>> odesys = SymbolicSys.from_callback(f, 2, 1)\n >>> xout, yout, info = odesys.integrate(10, [1, 0], [1], integrator='odeint', nsteps=1000)\n >>> _ = odesys.plot_result()\n >>> import matplotlib.pyplot as plt; plt.show() # doctest: +SKIP\n\n.. image:: https://raw.githubusercontent.com/bjodah/pyodesys/master/examples/van_der_pol.png\n\nIf the expression contains transcendental functions you will need to provide a ``backend`` keyword argument:\n\n.. code:: python\n\n >>> import math\n >>> def f(x, y, p, backend=math):\n ... return [backend.exp(-p[0]*y[0])] # analytic: y(x) := ln(kx + kc)/k\n ... \n >>> odesys = SymbolicSys.from_callback(f, 1, 1)\n >>> y0, k = -1, 3\n >>> xout, yout, info = odesys.integrate(5, [y0], [k], integrator='cvode', method='bdf')\n >>> _ = odesys.plot_result()\n >>> import matplotlib.pyplot as plt\n >>> import numpy as np\n >>> c = 1./k*math.exp(k*y0) # integration constant\n >>> _ = plt.plot(xout, np.log(k*(xout+c))/k, '--', linewidth=2, alpha=.5, label='analytic')\n >>> _ = plt.legend(loc='best'); plt.show() # doctest: +SKIP\n\n.. image:: https://raw.githubusercontent.com/bjodah/pyodesys/master/examples/lnx.png\n\nIf you already have symbolic expressions created using e.g. SymPy you can create your system from those:\n\n.. code:: python\n\n >>> import sympy as sp\n >>> t, u, v, k = sp.symbols('t u v k')\n >>> dudt = v\n >>> dvdt = -k*u # differential equations for a harmonic oscillator\n >>> odesys = SymbolicSys([(u, dudt), (v, dvdt)], t, [k])\n >>> result = odesys.integrate(7, {u: 2, v: 0}, {k: 3}, integrator='gsl', method='rk8pd', atol=1e-11, rtol=1e-12)\n >>> _ = plt.subplot(1, 2, 1)\n >>> _ = result.plot()\n >>> _ = plt.subplot(1, 2, 2)\n >>> _ = plt.plot(result.xout, 2*np.cos(result.xout*3**0.5) - result.yout[:, 0])\n >>> plt.show() # doctest: +SKIP\n\n.. image:: https://raw.githubusercontent.com/bjodah/pyodesys/master/examples/harmonic.png\n\nYou can also refer to the dependent variables by name instead of index:\n\n.. code:: python\n\n >>> odesys = SymbolicSys.from_callback(\n ... lambda t, y, p: {\n ... 'x': -p['a']*y['x'],\n ... 'y': -p['b']*y['y'] + p['a']*y['x'],\n ... 'z': p['b']*y['y']\n ... }, names='xyz', param_names='ab', dep_by_name=True, par_by_name=True)\n ... \n >>> t, ic, pars = [42, 43, 44], {'x': 7, 'y': 5, 'z': 3}, {'a': [11, 17, 19], 'b': 13}\n >>> for r, a in zip(odesys.integrate(t, ic, pars, integrator='cvode'), pars['a']):\n ... assert np.allclose(r.named_dep('x'), 7*np.exp(-a*(r.xout - r.xout[0])))\n ... print('%.2f ms ' % (r.info['time_cpu']*1e3)) # doctest: +SKIP\n ... \n 10.54 ms\n 11.55 ms\n 11.06 ms\n\nNote how we generated a list of results for each value of the parameter ``a``. When using a class\nfrom ``pyodesys.native.native_sys`` those integrations are run in separate threads (bag of tasks\nparallelism):\n\n.. code:: python\n\n >>> from pyodesys.native import native_sys\n >>> native = native_sys['cvode'].from_other(odesys)\n >>> for r, a in zip(native.integrate(t, ic, pars), pars['a']):\n ... assert np.allclose(r.named_dep('x'), 7*np.exp(-a*(r.xout - r.xout[0])))\n ... print('%.2f ms ' % (r.info['time_cpu']*1e3)) # doctest: +SKIP\n ... \n 0.42 ms\n 0.43 ms\n 0.42 ms\n\nFor this small example we see a 20x (serial) speedup by using native code. Bigger systems often see 100x speedup.\nSince the latter is run in parallel the (wall clock) time spent waiting for the results is in practice\nfurther reduced by a factor equal to the number of cores of your CPU (number of threads used is set by\nthe environment variable ``ANYODE_NUM_THREADS``).\n\nFor further examples, see `examples/ <https://github.com/bjodah/pyodesys/tree/master/examples>`_, and rendered\njupyter notebooks here: `<http://artifacts.bjodah.se/pyodesys/branches/master/examples>`_\n\nRun notebooks using binder\n~~~~~~~~~~~~~~~~~~~~~~~~~~\nUsing only a web-browser (and an internet connection) it is possible to explore the\nnotebooks here: (by the courtesy of the people behind mybinder)\n\n.. image:: http://mybinder.org/badge.svg\n :target: https://mybinder.org/v2/gh/bjodah/pyodesys/v0.11.6?filepath=index.ipynb\n :alt: Binder\n\n\nCiting\n------\nIf you make use of pyodesys in e.g. academic work you may cite the following peer-reviewed publication:\n\n.. image:: http://joss.theoj.org/papers/10.21105/joss.00490/status.svg\n :target: https://doi.org/10.21105/joss.00490\n :alt: Journal of Open Source Software DOI\n\nDepending on what underlying solver you are using you should also cite the appropriate paper\n(you can look at the list of references in the JOSS article). If you need to reference,\nin addition to the paper, a specific point version of pyodesys (for e.g. reproducibility)\nyou can get per-version DOIs from the zenodo archive:\n\n.. image:: https://zenodo.org/badge/43131469.svg\n :target: https://zenodo.org/badge/latestdoi/43131469\n :alt: Zenodo DOI\n\n\nLicenseing\n----------\nThe source code is Open Source and is released under the simplified 2-clause BSD license. See `LICENSE <LICENSE>`_ for further details.\n\nContributing\n------------\nContributors are welcome to suggest improvements at https://github.com/bjodah/pyodesys (see further details `here <CONTRIBUTORS.rst>`_).\n\nAuthor\n------\nOriginal author: Bj\u00f6rn I. Dahlgren (gmail address: bjodah).\nSee file `AUTHORS <AUTHORS>`_ for a list of all authors.\n",
"bugtrack_url": null,
"license": "BSD",
"summary": "Straightforward numerical integration of ODE systems from Python.",
"version": "0.14.3",
"project_urls": {
"Homepage": "https://github.com/bjodah/pyodesys"
},
"split_keywords": [],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "cadc15454bd8d120e10105b36a9321b0dd9ee15e2b1533f61d53261cd14399a6",
"md5": "0905451a33c2651adc572bf04c62f7e8",
"sha256": "7a5320dbcadc33c2d2b83c48393c03b4bd97f76c9c9fb075d8dbd28faddd19f6"
},
"downloads": -1,
"filename": "pyodesys-0.14.3.tar.gz",
"has_sig": false,
"md5_digest": "0905451a33c2651adc572bf04c62f7e8",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.7",
"size": 96359,
"upload_time": "2024-04-24T12:15:59",
"upload_time_iso_8601": "2024-04-24T12:15:59.163825Z",
"url": "https://files.pythonhosted.org/packages/ca/dc/15454bd8d120e10105b36a9321b0dd9ee15e2b1533f61d53261cd14399a6/pyodesys-0.14.3.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-04-24 12:15:59",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "bjodah",
"github_project": "pyodesys",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"circle": true,
"lcname": "pyodesys"
}
JSON
