This is primarily a test of using the researchblog.py script on Franklin. The results should be equivalent to what we saw when we did this for Cyclades64.
Also, it appears that we can’t stick the commands into a separate shell script here. We need to aprun each thing from within the jobfile that gets submitted to PBS.
These are from revision 268. The purpose of these runs is to test how (and if) we converge to the results reported in JCP 110, 4828 fig3 as we increase the number of modes and the width of the spectral density sampling window.
This is primarily a test of using the researchblog.py script on Cyclades64. However, these results are actually relevant, too.
RMSD at time 5
RMSD at time 10
RMSD at time 15
RMSD at time 20
RMSD at time 25
Movie showing how the error bars and averages converge. This is just over the first 40k trajectories (it gets pretty boring, visually, after that).
This is, in a way, a repeat of the plots from May 19 of a similar name.
By using the RMSD instead of the standard deviation, I think we account for any unfairness in the comparison due to sorting.
This plot compares the converged results of the Frenkel and Metropolis methods.
This plot shows the phase space covered by the two methods.
RMSD at time 5
RMSD at time 10
RMSD at time 15
RMSD at time 20
RMSD at time 25
Movie showing how the error bars and averages converge. This is just over the first 40k trajectories (it gets pretty boring, visually, after that).
These are new visualizations of the results from the Frenkel vs. Metropolis runs from May 18.
These new versions include RMSD instead of just standard deviations, as well as all sorts of pretty visualizations.
The analysis routines did not run last time, so we try again. These are based on the calculations from the previous post.
All of these are with the local harmonic prefactor. Obviously, the sorted Frenkel has an unfair advantage with few trajectories.
These might be a little bit unfair at small numbers of trajectories, since the Metropolis stuff we’re comparing to (calculated in previous run) blog is not sorted.
