Victor Lin

Seems to be the same as Hekstra-Lab/reciprocalspaceship#36 and https://groups.google.com/g/anvio/c/ayvtZoUWbJI but don't see any clear solutions there.

@jameshadfield downgrading to **2 - Usability w/ workaround** since the workaround is to use python=3.8.

Update: after experimenting with loading sequence data into the database (06e56981432be868939b9d40cadb0c529e99e38d), it turns out to be unnecessary overhead for `augur filter` and will come later. More info in [Slack thread](https://bedfordlab.slack.com/archives/C01LCTT7JNN/p1647029722758899).

@huddlej I like the idea of expanding `--metadata` to handle databases. Though we are postponing sequences in database for now, this will come up eventually and your point about having...

@huddlej > In this PR, I'm worried that modifying the sequence index directly and copying the metadata TSV to CSV will decouple the original source data from the test outputs....

> have the option of per-week sampling categories in addition to per-month sampling categories. I don't think `--group-by week` is right: - It's more difficult to extract that info from...

Oh no, sorry you had to find this out the hard way. I think raising a custom error message would be better than this cryptic error message that comes with...

I haven't tried with [auspice/CHANGELOG.md](https://github.com/nextstrain/auspice/blob/master/CHANGELOG.md?plain=1) yet. But from a quick look, it seems to follow the same format of `## ` header per version, which would work.

Rebased onto #997 so 13f46d49b5816574ee34e2065e99791933532c97 is the only change to review here. I can't update the GitHub base branch to [`victorlin:filter/split-file`](https://github.com/victorlin/augur/tree/filter/split-file) since it's from a fork.

Will rebase/merge this after #997 goes through.