Thomas Duvernay

Thanks for having a look @ZwFink. The error message provided results from a `pip` installation of Charm4Py. I have just tried with an MPI build of Charm++ (`./build charm4py mpi-linux-x86_64...

@tylerjereddy @tupui Any chance for some feedback on this PR? Could you potentially add the proper tags (`scipy.integrate` and `defect` I imagine) and request a review from someone familiar with...

Thanks @tylerjereddy. I agree I would also like to provide the proper explanation as to why this behaviour is happening. Having a look again at the last update I posted...

@tylerjereddy I have just found the following [comment](https://github.com/scipy/scipy/blob/774941b24a10593dfb070f6dfdb702e0fc2fffce/scipy/integrate/_ivp/tests/test_ivp.py#L204-L208), introduced within this [commit](https://github.com/scipy/scipy/commit/7e2ddf86c5f8df9642f330e4a1d09f838e953306) by @nmayorov, which corresponds exactly to the bug I have reported. I will push a commit to remove...

It turns out that these tests were passing anyway even without the changes proposed in this PR. I have tried to further replicate the behaviour I have previously described, but...

@tupui Thanks for the heads-up! CI is back to green, so I can give more detail about what happened over the last few commits. Reverting the changes in `lsoda.py` and...

_Update regarding the last commits_ Through a short discussion with the developer of the Julia wrapper around the C implementation of LSODA, I realised that the way I was describing...

@nmayorov Would you have some time to take a look at this PR? It would be nice to have this bug fixed.

Just a small comment that #15860 seems to be fixed by the changes made to `lsoda.py` in this PR.

@mdhaber I have rewritten the commit history of this PR and now have only two commits. The first one introduces changes in `OdeSolution`, re-activates testing of LSODA in some tests,...