Software! Math! Data! The blog of R. Sean Bowman
The blog of R. Sean Bowman

Posts with tag “d3”

November 11 2016

In distributed systems a process often needs to wait for more than one of its peers before completing some action. For example, we may send a request to 5 of our peers and require that we hear back from 3 of them before considering the action completed. Even when the distribution of round trip times for a single request is pretty simple, it’s not entirely clear what the distribution of the aggregate request will be. (To me, at least; maybe if you’re smart or know statistics well it’s simpler.) In any case, we can use a histogram to get an idea of the shape of the density function.


August 13 2016

D3 is a Javascript library for creating charts, animations, and other visualizations. It’s awesome. One of its key features is the easy creation of a DOM heirarchy mirroring some data you’ve carefully constructed for the job. Here’s a way to create a hierarchy of elements where the elements in each level aren’t necessarily homogeneous. I used this to make a table with a header column, but I’m sure it has lots of uses.


February 11 2016

In a previous article, we saw how to compute the (zeroth) persistent homology of a filtered simplicial complex. But how do we obtain such a complex, anyway? Scientists and others who work with data often start with a bunch of points in a high dimensional space. What are some good ways to turn that information into a filtered simplicial complex we can compute the persistent homology from?


February 09 2016

I’d like to introduce an algorithm for finding clusters in graphs. This is an important problem in machine learning, in part because so many problems can be phrased as finding “nice” clusters in a given graph, where the meaning of "nice" tends to depend on the particular application. It’s easy to imagine simple applications of a good clustering algorithm, like finding groups of friends in a social network. But things only get more interesting from there.