This weekend I’m participating in Yes!Delft’s 1-2-Launch weekend startup program. It’s been an absolutely blast and I’m completely psyched about the business we’re launching!

It’s been a great experience getting out on the street and figuring out what problems people are facing and want solved.

Enter GoUrbanGrow, your houseplant saviour! Our vision is a world in which houseplants don’t die for dumb reasons. Our mission, should you choose to accept us, is to help you save them :).

Anyway, tomorrow we are pitching to the 1-2-Launch jury and the winner (there’s always a winner!) is given the opportunity to take part in Yes!Delft’s brilliant Launchlab program.

We want to win of course!

So if you’re intrigued, visit us at http://gourbangrow.com and leave us your email address to learn more!

Thanks for listening 🙂

Meant to be


When nature blossoms,
Earth expresses love,
intangible yet recognisable.

The relentless rustling of trees,
trickling of water between rocks,
exposes harmony and melody,
provides glimpses of perfection.

This idyllic beauty,
is only surpassed by the radiance,
you bring to my life.

Love so strong, it wilts the will of all that oppose it.
Love so deep, it takes me to places beyond the physical realm.
Love so commanding, it captures my every sense.

I take a moment to reflect,
and it reaffirms my faith,
that you and I are meant to be.

Spyder — MATLAB for Python


Having “grown up” on MathWorks MATLAB for my high-level scientific computing applications, I’ve grown accustomed to the easy, click-to-inspect interface that MATLAB provides when prototyping new algorithms. MATLAB is sold as the Language of Technical Computing, with good reason. For rapid prototyping and visualization, it’s a great package, and the folks at MathWorks are continuously improving the interface.

The major downside however is that with great quality comes great prices (or something like that …). I’m still enjoying being able to use an academic license at the moment, however, not knowing where my future lies, I’ve realised that it’s probably a worthwhile endeavour to expand my horizons, break the shackles, and move to a free, open-source alternative.

Enter Python. I’ve been toying with the idea of stepping over to Python for my scientific computing applications for a number of years now, but have mostly held back because of my perception of a lack of maturity of scientific packages. Recently, I’ve discovered that this perception is completely unfounded. So, I’ve been gradually moving my scientific computing applications in MATLAB over to Python, using the PyLab libraries, and so far so good.

IPython is a great invention, and I’ve had the fortune of sitting in on a demo-based class given by the guru himself, Fernando Perez, at UC Berkeley (unfortunately, I had to leave Berkeley before being able to attend one of the Python bootcamps that I’ve heard so much about). The class convinced me that Python is mature enough now to really consider giving up on my MATLAB addiction.

I’ve tried using IPython Notebook for a while now, and as much as I love the /-like worksheet approach, it’s definitely too cumbersome to prototype code. For all my code-editing purposes, I’m using Sublime Text 2 at the moment (if you’re not on the bandwagon yet, what da heck are you waiting for???), however, I miss the ability to inspect the value of variables, objects, structures by a simple click. In addition, MATLAB’s debug and profiling utilities are fantastic, and I’ve only realised since moving to Python how much I make use of them.

So, struggling to figure out whether I can stick to Python for all my high-level code development, I stumbled across Spyder. My first day or so of coding within the Spyder environment is great … feels like I never left MATLAB! Sure, there are a number of differences, and in particularly, I’m not a fan yet of the editor, which I hope to be able to tinker with (if anyone knows if there’s a way to plug the ST2 editor in Spyder, hit me up!). In any case, I’m genuinely happy, and not missing MATLAB for one second!

At the moment, my only remaining gripe with regards to the Python-MATLAB substitution is that I REALLY don’t like the 3D plotting capabilities in Python … plots look well, downright awful. MATLAB is really nice in this regard. If I can solve this, then I think I can say GOODBYE MATLAB … something to look into …