search

msmbuilder / msmbuilder-legacy

Legacy release of MSMBuilder
http://msmbuilder.org
GNU General Public License v2.0
25 stars 28 forks source link

Confusing percentages printed in BuildMSM.py #404

Open dvanatta opened 10 years ago

dvanatta commented 10 years ago

BuildMSM.py prints: 14:00:58 - Selected component 387 with population 0.993999 . . . 14:02:45 - Ergodic trimming discarded: 0.589649 percent of your data

The first number is the raw population and the second number is already multiplied by 100.

I understand this is entirely human error on my part but perhaps it would be more clear if the first number was also a percentage?

Serilin commented 10 years ago

Hi, dvanatta Recently, i have been studying the MSM. And I can't understand the " Selected component 387 with population 0.993999". Does it mean that equilibration population of 387th cluster is 0.993999. If so, this population may be too large. Could you kindly give some explaination? Sorry for my innocent question, and thank you very much.

Best Serilin

dvanatta commented 10 years ago

Hi Serilin,

No, this is not the equilibrium population of the 387th cluster.

This step in the MSM creation process is where we do ergodic trimming to find the most connected subgraph because the state transition matrix can become disconnected at longer lagtimes.

The print statement above is telling us that component 387 (not exactly a cluster but this index is arbitrary) is connected to 99.3999% of my data. In other words, my state transition matrix is still well connected at the lag time in question.

-Dan

Serilin commented 10 years ago

Dan,

Thanks for your clear explanation, and I see.

Seriline