Daniel Weindl
Daniel Weindl
Caching remote files for tests (#2313) works well where it's used, but it's missing from `test_windows.yml` and `test_python_ver_matrix.yml`. Cache the [pooch directories](https://github.com/fatiando/pooch/blob/d33a98997e0f937f8fe7e3e20b3b47419cd07c9a/pooch/utils.py#L73) there. https://github.com/dweindl/AMICI/actions/runs/8259666961/workflow https://github.com/dweindl/AMICI/actions/runs/8259666780/workflow
Get rid of https://github.com/AMICI-dev/AMICI/blob/aa39743e794837461fe3286d52b5257f26ddee6b/python/sdist/amici/swig_wrappers.py#L131 This prevents using e.g. `isinstance(x, amici.ExpData)` (only works via `isinstance(x, amici.amici.ExpData)`). The same functionality can easily be injected into the swig-generated wrappers.
`pip3 install -ve git+https://github.com/AMICI-dev/amici.git@develop#egg=amici\&subdirectory=python/sdist` seems to fail on Windows - see https://github.com/AMICI-dev/AMICI/issues/2151#issuecomment-1655695489 Can any other Windows user reproduce that? (Note: This might be more of a symlink issue than an...
Currently this produces confusing warnings: > 2024-03-07 15:24:58.603 - amici.swig_wrappers - WARNING - [AMICI:Inf] AMICI encountered a Inf value for ts[0] (0) at t=35.802095 A check for the current `t`...
Add example for Solver.setContstraints and Model.setStateIsNonNegative
Support for AlgebraicRules was added in #2017. Sensitivity analysis is not yet supported. Looking for volunteers to finish that. Test case: `pytest tests/testSBMLSuite.py::test_sbml_testsuite_case[00039]` (and many others in this suite)
Currently, ReturnData has a number of protected data members that are used as temporary storage for model quantities during collecting the output data. It's currently rather intransparent which values are...
The CVODES constraint handling (see also #2327) would also allow us to enforce conservation laws, and probably make our code quite a bit simpler. To be discussed whether we'd want...
I am not sure if the current handling of conservation laws for non-constant species is correct for models that also contain constant species. I didn't check that in detail, but...