Andrew Pontzen

There seem to be two related issues about generating valid URLs for this database. I've tried fixing them in pull request #158 - could you see if this fixes your...

You could force this by implementing a custom handler I think -- take a look here https://pynbody.github.io/tangos/custom_input_handlers.html In your custom input handler, you'd want to override `get_stat_file` and force it...

Yes, I do on balance think it would be a good idea for the pre-implemented input handlers to select a particular stat file if that is what they are expecting....

I've tried a fix for this in pull request #120 (branch `issue-117`). See the notes there for how to use. Let me know if this helps @trquinn @mtremmel.

It has to be generated at least for the whole major progenitor branch, otherwise it’s not defined. But it would be nice if a neater/more helpful error resulted...

Ah, I see, yes it would be useful to implement a simple test to catch this type of inconsistency

It looks like the halo catalogue hasn’t implemented `get_group_array` for individual families. I assume this is using the AdaptaHOP catalogues? See https://github.com/pynbody/pynbody/blob/master/pynbody/halo/__init__.py#L208 for how it is handled with regular HOP.

I confirm this is too challenging to fix for now as it relates to numpy internals changing between versions.

Yes that's almost certainly a filesystem bug (although goes slightly in the opposite direction to what you'd expect, but I have seen weirder). As ever, we should not be using...

I think there's some mpirun / mpiexec variant that will label which line came from which processor?