Erik Bernhardsson

But will that work when the accuracy is exactly 1 or 0? Won't those be projected to +-infinity? Do you have a screenshot you can show?

I think something that warps [0, 1] using roughly `x^10` might also be OK. That would project 0.9 to 0.35 and 0.99 to 0.9 Not sure how hard it is...

Nice! I think we might want to throw in minor gridlines, but looks pretty good!!

Nice, I think this looks really great. Happy to merge!

i'm fine using the separate package instead of sklearn

I haven't seen that one but it looks cool! If you're interested in getting it added to ann-benchmarks, feel free to put together a pull request!

great if anyone wants to add it!

might be interesting to add a very low dimensional dataset (3-10 dims) for benchmarks

one idea is to just take some free data set on world cities lat/lon and project it into 3d space. i have some code for this lying around already.

That makes sense. It would be great if there's already some kind of popular dataset that we can rely on – I don't want to make it too "arbitrary". If...