germanjoey

In my experience, "Google Instant Answers" are more often than not factually incorrect. It makes me queasy every time I see that... it speaks volumes that's so much more important to them that they keep you on the page that they'd rather give you wrong information than let you go.

He's talking about Hunter S. Thompson in that passage you quoted, not himself.

Is this the reason why top recipe results always contain the poster's entire life story before they ever show a list of ingredients?

And here I thought the article was going to be like this parody manga about Alien Mahjong...... https://mangadex.org/chapter/56995/1

He likely wants to decrease the letter's contents' relevancy ranking in a Google search for "Timnit Gebru".

What's "Brain Women and Allies"?

Seems extremely similar to Appsheet... what's your differentiator here?

I am very interested in a version that allow you to pass in a custom substitution cost! It has a great use for OCR applications!

What do you think about a DSL alternative?

Looks awesome. I've used FuzzyWuzzy to some success in the past but its performance was abysmal, so I'm interested in giving this a try.

Any chance you could also implement a weighted-levenshtein version? (or allow passing in a weight-table for various characters combinations as like a kwarg or something?)

edit: Something like this, but with the speed of C++ would be amazing. https://pypi.org/project/weighted-levenshtein/

Audio starts around 20s.

Yep. Read about it here: http://shameproject.com/profile/malcolm-gladwell-2/

Hey, congratulations, this is a really cool algorithm! Thanks for sharing it.

I'm interested in this paper because I worked on a somewhat related problem some time ago, but got stuck on how to handle data that morphs into a mixed-modal distribution. Modes that are close together are no big deal, but modes that are spaced more exponentially apart are tricky to deal with. For an example of something that would be in DataDog's purview, it would be like trying to sketch the histogram of response times from an endpoint that sometimes took a "fast" path (e.g. a request for a query whose result was cached), sometimes took a "normal" path, and sometimes took a "slow" path. (e.g. a query with a flag that requested additional details to be computed) If the response times from the slow path is much bigger than the others, e.g. by an order of magnitude, their statistics might essentially drown-out the data from the other two paths since you're using them to calculate bin size.

I noticed you had some results from measuring DDSketch's performance on a mixed-modal distribution that looked pretty good (that "power" distribution on the last page). I was wondering if you had done any more investigation in this area? E.g. how messy/mixed can the data be before the sketch starts to break down?