Aggregate Knowledge Blog
blog.aggregateknowledge.com · Company blog · on HN since 2011
Engineering blog of Aggregate Knowledge covering probabilistic data structures, hash functions, streaming algorithms, and big data infrastructure.
What HN saysCommenters praise its HyperLogLog post as a very good explanation and recommend its hash function series and the interactive visualization.
15HN people shared it
7linked it in comments
12stories submitted
217story points
13 years agolast shared
2011people per year2026
Most-shared posts
- 1Sketch of the Day: HyperLogLog — Cornerstone of a Big Data Infrastructure6 people · 11 points · first shared 2012 · discussion on HN
- 2Choosing a Good Hash Function3 people · 2 points · first shared 2012 · discussion on HN
- 3AWS Redshift: How Amazon Changed The Game2 people · 187 points · first shared 2013 · 21 comments on HN
- 4Adventures in Concurrency1 people · 4 points · first shared 2012 · discussion on HN
- 5Streaming Algorithms and Sketches1 people · 4 points · first shared 2011 · discussion on HN
- 6Open Source Release: postgresql-hll (postgres hyperloglog)1 people · 3 points · first shared 2013 · discussion on HN
- 7Big Memory, Part 31 people · 2 points · first shared 2012 · discussion on HN
- 8Choosing a Good Hash Function, Part 21 people · 2 points · first shared 2012 · discussion on HN
- 9Postgres HLL Intersections1 people · 1 points · first shared 2013 · discussion on HN
- 10HyperLogLog Venn diagram in JavaScript1 people · 1 points · first shared 2013 · discussion on HN
- 11Sketch of the day frugal streaming1 people · first shared 2013
- 12Choosing a good hash function part 21 people · first shared 2012
- 13Choosing a good hash function part 11 people · first shared 2012
What HN people wrote when they linked it
nefasti · 2014-04-01
A very good post helping explain HyperLogLog : http://blog.aggregateknowledge.com/2012/10/25/sketch-of-the-day-hyperl…
jacinda · 2014-04-01
The top answer about HyperLogLog on stackoverflow links to this awesome visualization: http://www.aggregateknowledge.com/science/blog/hll.html
SO Answer: http://stackoverflow.com/a/13074072/307293
Full blog post: http://blog.aggregateknowledge.com/2012/10/25/sketch-of-the-day-hyperl…
abolibibelot · 2013-10-29
According to this comment http://blog.aggregateknowledge.com/2012/10/25/sketch-of-the-day-hyperl… , it's custom made and based on d3.
noelwelsh · 2013-10-14
Nice analysis.
Since that post was submitted Murmurhash 3 and a number of other notable hash functions have been released. If you liked the linked post you should go read Aggregate Knowledge's series of posts on hash functions for a look at some newer algorithms: http://blog.aggregateknowledge.com/2012/02/02/choosing-a-good-hash-fun…
pnathan · 2013-09-18
Here's an interesting probabilistic way to sample up to the median:
http://blog.aggregateknowledge.com/2013/09/16/sketch-of-the-day-frugal…
gizzlon · 2012-05-29
Found these by coincidence, might be interesting to some of you.. (can't speak to the content)
http://blog.aggregateknowledge.com/2011/12/05/choosing-a-good-hash-fun…
http://blog.aggregateknowledge.com/2011/12/29/choosing-a-good-hash-fun…
http://blog.aggregateknowledge.com/2012/02/02/choosing-a-good-hash-fun…
ricardobeat · 2013-05-18
Use wix.com[1], learn to code[2]. In your description you say you have set up an OpenStack install at home, HTML/CSS is a walk in the park compared to that. Use launchrock[3] or mailchimp[4] to collect e-mails.
"big data" really means big - as in millions of dollars big, tera/petabyte scale. Not going to happen with $5k. You will suffer to get anything close to AWS prices[5]. You could brand the service as "fat" servers or something similar until you can really tackle that kind of scale.
And, honestly, you should omit the thing about meeting Calvin Klein and DJs, it doesn't leave a good impression. Tell us what you have to offer as a founder, and get to work on a business plan with real cost figures.
[1] http://wix.com
[2] http://www.codecademy.com/courses/html-one-o-one
[3] http://launchrock.co
[4] http://mailchimp.com
[5] http://blog.aggregateknowledge.com/2013/05/16/aws-redshift-how-amazon-…