Page 1 of 1

Problems with CalculateImpliedTimescales.py

Posted: Tue Nov 27, 2012 3:43 am
by bearshare
Dear all:
I'm new with the MSMbuilder, while I followed the tutorial of alanine dipeptide without any problem, I encounter the problem of "Exception: you cannot calculate 11 eigenvectors from a 2X2 matrix" with my own testing trajectory .

My input files are 50 10ns trajectories using AMBER(10000frames, using the methods of targeted MD), which have been converted into XTC format using catdcd.

Firstly, I used the command: ConverDataToHDF.py -s mypdb.pdb -i XTC,then : Cluster.py rmsd hybrid -k 1000. After that I used the command: CalcylateImpliedTimesscales.py -l 1,25 -i 1 -o Data/ImpliedTimescales.dat, but it gave me an error, "Exception:you cannot calculate 11 eigenvectors from a 2X2 matrix".

I tried many times , but still can not solve the problem, I think the problem might be wihtin the second step when clustering into microstates, but i cannot fix it, could anyone tell me what is the problem and how can I fix it?

Also, i am confused with Cluster.py, what is the difference between "-d " and "-k ", also what is the suitable value for -d and -k?

Re: Problems with CalculateImpliedTimescales.py

Posted: Wed Feb 27, 2013 3:14 pm
by rmcgibbo
Shan,

That error message means that the model produced only has two states, and as such cannot have 10 timescales -- each timescale corresponds to an eigenvector of the transition matrix and a 2x2 matrix can only have 2 linearly independent eigenvectors. This is probably being caused by the ergodic trimming step. I would try clustering into fewer states than 1000 initially.

Also, sorry for not checking this forum for a very long time. Please sign up for the msmbuilder-user email list, which will hopefully be a little more active than these forums. https://mailman.stanford.edu/mailman/li ... ilder-user

-Robert