msultan

The ```trimmed_assignments``` is necessary if your MSM throws away some data. The MSM re-indexes the states internally and the trimmed assignments might become necessary if the strongly connected component is...

i wonder if we need to push all the features through a Normalizer before we can use some of these algorithms. We should be careful about that. The variance threshold...

to be honest, something naive would be nice to get up and running. my thinking is that the "human supervision" is good at finding the crossover because we can abstract...

>how can I obtain the representative structures of the microstates or macrostates of the MSM that I built ? Please take a look at the following ipython notebook which walks...

That really shouldnt be happening since we dump and then reload the same txx data in the second ipython notebook. Two things come to mind: 1). Did you sub-sample the...

i think your ```ttrajs``` is not the right shape. ```ttrajs```should be of the shape N,T where N is the number of frames and T is number of tICs. Could you...

Also thank you @jeiros , I wouldn't have been able to explain it better myself. You are right, the tICA-trajectories have no time information, they are designed to simply try...

>As for the trajectories based on MSM I wonder what's the time order for them? The time between frames is the MSM lagtime. >Since I notice that some structure snapshots...

any particular reason for the change? i honestly dont know if it matters either way.

Some thing else to consider http://journals.plos.org/ploscompbiol/article/file?id=10.1371/journal.pcbi.1004568&type=printable