On the side I’ve built a few (mostly open-source) projects, partly to learn new technologies, and partly because I find the topic interesting. I’m not actively working on more than one at a time, but I keep them all working (I’m happy to respond to questions, bugs, or suggestions):
bunker.land is a project which maps out where in the US you don’t want to live during (1) nuclear war or (2) natural disasters. Basically I’m try to recreate the old cold war nuclear target maps, plus other natural disaster layers. The aggregate maps look like this:
jbool_expressions is a lightweight library for building and manipulating dynamic boolean expressions in Java. Among other things, it can take an arbitrary boolean logical statement:
and heuristically simplify it a minimal equivalent:
(A & B)
It also supports conversion to Disjunctive Normal Form and Conjunctive Normal Form.
nlpviz is a simple visualizer which uses the stanford-corenlp library to parse arbitrary English (or Spanish) sentences and display them in a tree structure using d3 + dagre. It displays sentences like this:
I run a copy of the site here.
taxtree is a visualization of the taxonomic tree of all known species, using the Catalog of Life index. It also uses d3 + dagre to display species as a tree:
I run a copy of it here.
Uncharted is an explorable in-browser map of earth’s solar neighborhood (all stars within 75 light years, plus known exoplanets), using three.js. I do my best to accurately represent colors and distances:
You can explore it here.
LiveRamp / Big Data tools
I helped built a lot of Big-Data tools at LiveRamp, and several of them are open-source. I’m no longer involved in developing these, but I’m happy to respond to PRs or questions:
Workflow2 is a DAG processing engine LiveRamp uses to help engineers quickly build failure-resilient, high performance, complex data processing pipelines. Check out the repository’s README for examples and features.
The cascading_ext project is a collection of tools built on top of the Cascading platform which make it easy to write high-performance Hadoop workflows. Some posts describing tools available in the library: