Eric Larson

Do you need a true median or is an approximation sufficient? There are some cool approximations here https://stackoverflow.com/questions/638030/how-to-calculate-or-approximate-the-median-of-a-list-without-storing-the-list

in particular, you could do, say, an 8-fold `remedian` calculation, and with `n_jobs=8` it would give an 8-fold speedup (in theory)

Maybe just `EpochsArray` with `np.random.rand` and `create_info`?

Not sure if this is the same thing, but there is no way to control verbosity. The `verbose` decorator from MNE could be ported or used direcly.

Not here: https://github.com/autoreject/autoreject/blob/master/autoreject/autoreject.py#L140 And it uses `print`

IIRC in MNE we use `print` only for progressbars (because they delete) and `verbose` for everything else because it routes through `logger`, which gives lots of advantages (e.g., standard file-output...

We also do something like this in MNE. With the `verbose` decorator, you just see if `logger.level

Yes 18 digits is fine for that test, feel free to open a PR Indeed we should dig into whether it's an OpenBLAS bug or if we need to adjust...

> The emulation is not fast, but it works. Okay I'll give this a shot. I'm on Ubuntu 22.04 but I'm assuming the commands are easy to translate. I'll post...

> you can install qemu-efi and then use virt-manager to set up an emulated VM using the appropriate ISO. I set up a qemu image using ppc64le arch and a...