pyIPCA a python library for Incremental PCA

I extracted some of the useful code and nifty examples from the background of my Thesis as a python library for your enjoyment. PCA or Principal Component Analysis is a pretty common data analysis technique, incremental PCA lets you perform the same type of analysis but uses the input data one sample at a time rather than all at once.

The code fully conforms to the scikit-learn api and you should be able to easily use it anywhere you are currently using one of the sklearn.decomposition classes. In fact this library is sort of on the waiting list for sklearn.

IPCA on 2D point cloud shaped like an ellipse: figure

Check it out if you’re interested and holla at sklearn if you want this feature!

read more

Weekend Project - Install SteamOS

This weekend I championed my way through installing SteamOS (the Debian distro by Valve that will be the installed on the upcoming Steam boxes). I had to do some pretty crazy stuff to get it working including dropping out of the automated install to manually inject grub-pc and then compiling the drivers for my wireless card. All in all it was a triumph!

and then finally:

This was an early beta release but they made some weird choices - like handicapping the basic Debian installer by fully automating it and only supporting efi. I was actually a bit disappointed when I finally finished because the end result is not really different from simply installing Ubuntu and setting Steam big picture mode to auto start, I am not sure what exactly I was expecting though. SteamOS is much more for OEMs than the DIY crowd at the moment but I can see that Valve is super invested in Linux at this point with a ton of additions to their own repositories. Good things are going to come of this I can feel it!

* Edit *

Almost all the hacking I had to do has been wrapped in Ye Olde SteamOSe 

* Edit 2 *

Wow Valve released an updated version of the beta addressing a lot of the problems Ye Olde SteamOSe addressed and they allegedly collaborated to get this done! This is why Valve is going to win the next generation - working with the community. Full story here

read more

Book Review: Android Application Programming with OpenCV

I was asked to do another book review for Packt Publishing which I announced a little while ago. I’ve been pretty busy with my new job (which is awesome! I’ll probably make a post about it and all the stuff I’ve learned soon!) but I eventually finished giving the book a quick read through.

It was a good book! I wish I’d had the time to follow along and make my own app but unfortunately I didn’t have time. The book has lots of detail and screenshots to guide you through making your own app which is helpful for working with more graphical type things like Eclipse. I think this book would save a developer looking to get started with Android and OpenCV a lot of time and remember people time is money (money you could spend on this book). Reading the book really made me want to start an Android OpenCV project, I think the mobility of the platform makes it so much more fun! To think all my computer vision projects have been tied to a desktop or a best a laptop until now is such a shame.

I think I like this model of offering premium lessons and documentation at a price and as long as it stays clearly on the premium side and not “a pay to even get docs” situation I think it’s a good way to get some money back into Open Source and the people who make it happen. Packt does it right too paying some of the highest royalties of any publisher of IT books.

tl;dr If you’re planning on doing some OpenCV Android development buy this book - it will undoubtedly pay for itself in time savings

You can get the book here:

read more