• Do refactors quickly in complete thin layers

    Warning. This is unabashedly opinionated and doesn’t provide any references or hard evidence, but it is hopefully sensible advice nevertheless. The pitch Current conventional wisdom in software says that you should tie all work to externally visible value delivery, and that if you don’t need to touch something it’s better left alone. That’s wrong. If…

  • Who needs distributed systems anyway?

    I spent the last 2-3 months building a system to synchronize BizOps data between various third party products. It was an interesting challenge to build something so complicated in such a short space of time. I opted for an extremely flexible monolithic design, with an internal “micro event loop” that operates on batches of data.…

  • Scaling Map Rendering with Mapbox & Google [crosspost]

    Soon after joining LandInsight I tackled an interesting scaling problem to do with rendering maps in the browser.  In a post on the company’s technical blog I explain more (you can find the project open sourced here).   How we Scaled our Mapping Tool with Mapbox & Google At the heart of the LandInsight product…

  • From neuroscience to software engineering [crosspost]

    Originally posted on GSN Munich: Dr. Daniel Manson is a neuroscientist-turned-developer based in London. At the moment, he is a software engineer at the London-based  LandInsight – a company providing services for land developers in England and Wales. But back in the day, he was a Neuroscience PhD student at the University College London. There he…

  • My new boss did an interview with me [crosspost]

    Land insight, my new employer, is a startup that provides property developers with an interactive mapping tool to explore various land and property datasets.  I’m on the software team, working as a full stack developer and machine learning evangelist. The following post, by CEO Jonny, was originally posted here, on the Land Insight blog. We thought…

  • Waveform GUI: neuroscience in Chrome

    During my PhD I used bundles of electrodes to record the voltage fluctuations generated by neurons in the brain.   I’m not going to describe the details of the science here, instead I want to focus on the graphical user interface I ended up building. As of this afternoon, I now have a video demo…

  • Should I be using pandas?

    My PhD consisted partly of doing very hands-on neuroscience experiments, and partly of analyzing the “medium-sized data” sets produced by those experiements. I say “medium-sized” because the data was large enough to require care, but not so large that it couldn’t fit comfortably on a single disk in raw form, and fit in memory once partially processed.…

  • Flags of the world (indexing tutorial)

    The code examples for this tutorial are currently only given for Python/Numpy. Conversion to Matlab is not difficult, but would be confusing for beginners (who are the target audience here). I may get round to doing the extra work at some point soon. This tutorial is aimed at absolute beginners. Ideally get an experienced person…

  • Non-trivial vectorizations

    For a more general discussion on optimizing, see my previous post on perfectly timed optimization. Before we get started, I should say that the following stories are likely to be of most interest to Matlab or Numpy users.  In compiled languages, the general lack of overheads will probably do a pretty good job of making fast programs, as…

  • Perfectly-timed optimization

    I was going to write about specific vectorization tricks in Numpy and Matlab, but this is what I ended up with instead.  Stay tuned for the actual vectorization stuff. Neuroscientists, and possibly biologists more generally, seem to take great pride in producing slow code.  They may even quote you Knuth’s famous line: …premature optimization is the root…

Create your own website, free on WordPress.com.Create your own website.
Start free