Example PCA implementation and analysis

pca.ipynb

This would make a nice tutorial!

For information: The trajectories can be obtained from Data download for the 2015 CECAM Workshop.

@kain88-de, do you want to link to this gist from https://github.com/MDAnalysis/mdanalysis/wiki/Tutorials ?

Btw, you could also color the residue B-factors by the magnitude of the variance in the 1st PC and display: This should show very clearly which domains are involved in the motion.

This is great, just posted about it on the google groups. Only possible comment that I'd have is that we might want to use an interactive plotting package like Plotly or Bokeh. Being able to mouseover specific points in the plotted data is nice.

@orbeckst We can make this a tutorial once @jdetle has implemented a PCA :-). I don't know much about the coloring in nglview.

@jdetle. replace %matplotlib inline with %matplotlib nbagg to get interactive plots. I personally like the static pictures here because they are rendered in github, the js isn't.