SVD is so cool!

I've been looking at using SVD (Singular Value Decomposition) to understand matrix components and analyse error. I had a meeting with Peter yesterday and he taught me more on it and got me very interested into really undertand what SVD really is.
So I search on the web and had a read on Singular Value Decomposition by Todd Will and watched a video lecture by Gilbert Strang in MIT. I also found some Matlab teaching codes that may be useful to help me fully appreciate some basic maths.

SVD is so useful! It's a good friend, not a threat! :)