Kyle Beauchamp

~~We probably also want an easy way / tutorial to calculate the tICA projections of each trajectory.~~ Edit: moved this request to a separate issue, as it's not docs related.

Also: need to cite your paper in the tutorial. Probably would also be nice to cite Frank's tICA paper as well.

OK looks like Robert has cited you in the HTML tICA theory guide, so that's already done.

Ok we also need to modify the commands to due to the changed tica load inteface.

Maybe if there is a boring talk at BPS next week On Feb 12, 2014 9:57 PM, "Robert McGibbon" [email protected] wrote: > Do you want to open a PR? >...

FWIW, I think the sklearn theme does a better job on this: http://scikit-learn.org/stable/modules/generated/sklearn.cluster.KMeans.html#sklearn.cluster.KMeans

Here's another example of how our theme is "bad": http://mdtraj.org/latest/api/generated/mdtraj.utils.ensure_type.html#mdtraj.utils.ensure_type

Here is the driver code: ``` import pymc import numpy as np from msmbuilder import mcmc num_states = 3 alpha = 1 C = np.ones((num_states,num_states)) C[0,0] = 1000 sampler =...

The reason I use the `iterate_K()` function is that we have to handle two cases: 1. The samples all fit in memory. 2. The samples are too big for in-memory

I'll add more comments and docstrings later. It might be useful for people to just ask questions about what isn't intuitive about the PyMC interface.