search

markovmodel / PyEMMA

🚂 Python API for Emma's Markov Model Algorithms 🚂
http://pyemma.org
GNU Lesser General Public License v3.0
306 stars 119 forks source link

NotImplementedError: MSM estimation with connectivity='none' is currently not implemented. #1591

Closed Harshitasahni closed 1 year ago

Harshitasahni commented 1 year ago

I am trying to build a msm model by using "pyemma.msm.estimate_markov_model(data_clust.dtrajs, lag =1, score_method='VAMP1', connectivity='none')"

If i don't have the connectivity set to 1, it uses "largest" with that i get another error when i do pcca, with nstates = 3, and active states = 3. I get the following error: IndexError: index 3 is out of bounds for axis 0 with size 3

thempel commented 1 year ago

Apparently your MSM transition matrix is not connected. Before using PCCA and playing with the connectivity parameter, I would carefully look at the implied timescales convergence. You can also compute MSMs with the function you're using above (using default connectivity), but using different lag times, and check out the resulting active set or active state fraction (msm.active_set or msm.active_state_fraction).

Harshitasahni commented 1 year ago

Thank you. I understood.

Harshitasahni commented 1 year ago

Thank you.

From: Tim @.> Sent: Wednesday, May 3, 2023 12:33 PM To: @.> Cc: Harshita @.>; @.> Subject: Re: [markovmodel/PyEMMA] NotImplementedError: MSM estimation with connectivity='none' is currently not implemented. (Issue #1591)

You don't often get email from @.*** Learn why this is importanthttps://aka.ms/LearnAboutSenderIdentification [EXTERNAL]

Apparently your MSM transition matrix is not connected. Before using PCCA and playing with the connectivity parameter, I would carefully look at the implied timescales convergence. You can also compute MSMs with the function you're using above (using default connectivity), but using different lag times, and check out the resulting active set or active state fraction (msm.active_set or msm.active_state_fraction ).

— Reply to this email directly, view it on GitHubhttps://github.com/markovmodel/PyEMMA/issues/1591#issuecomment-1533517248, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ANENNYH6ENFVYC7VVE365SDXEKQI7ANCNFSM6AAAAAAWLEG7QU. You are receiving this because you authored the thread.Message ID: @.***>