Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Discrete HMM initialization occasionally fails #25

Open
franknoe opened this issue May 12, 2015 · 0 comments
Open

Discrete HMM initialization occasionally fails #25

franknoe opened this issue May 12, 2015 · 0 comments
Assignees

Comments

@franknoe
Copy link
Contributor

For some initial transition matrices the discrete HMM initialization fails because the initial HMM has no reversible transition matrix. We need to enforce symmetry in order to remove numerical errors in calculating the hidden transition matrix from PCCA++.

Kathi's email:

Ich habe einen Testdurchlauf gemacht und dabei den Fehler bekommen, dass die versteckte transition matrix nicht reversibel ist.
In init -> discrete -> initial_model_discrete wird in 71-76 das besprochene Verfahren (paper) angewendet.
Kann es sein, dann nach
X = np.dot(np.diag(pcca.coarse_grained_stationary_probability), P_coarse)
das Symmetrisieren fehlt?
ich habe X = .5*(X+X.T) eingefügt und dann läuft es durch.

@franknoe franknoe self-assigned this May 12, 2015
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant