Closed BoundaryValueProblems closed 1 year ago
Thanks for the check. I really need to figure out a better way of testing these, the current method is too sensitive to the ~0 imaginary part. The error is only in the construction of the tests, though.
Just to let you know, I've been looking at this, and it turns out comparing floating points that have been pretty printed is actually a pretty thorny problem. Probably best to let this error slide by on either of our local copies or the mac until there's actual floating point comparisons done in doctests
Documenter.jl (if you're interested in more detail, see the mention from last week).
I've rewritten the tests so that the doctests only happen in the CI, so we'll only see meaningful local errors.
I upgraded Julia from v1.8.5 to v1.9 yesterday. And I just ran test of ContinuousWavelets.jl and got the error related to Documenter as follows. If you could check what needs to be changed and fix this, I highly appreciate it.