search

markovmodel / PyEMMA

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

No possibility to understand what are the clusters considered for HMM #1463

Closed pl992 closed 4 years ago

pl992 commented 4 years ago

I've checked deeply the documentation but I couldn't find any method that gives the possibility to get the active sets used for building the metastable states in a HMM. I'll try to be clearer: Starting from discrete trajectories (discretized by let's say 1000 clusters) I built a HMM. This HMM considers only a subset of the previous set (so let's say only 500 clusters). I would like to know by which clusters are made the metastable sets but since I don't have the information of which clusters are used (which are the active sets) I cannot use the output of hmsm.metastable_sets (I cannot relate the numbers provided to the cluster numbers I built). If this function isn't present it would be great having this information, if it is present then I would report that from the documentation isn't clear how to get this information. Thank you

thempel commented 4 years ago

Are you looking for hmsm.observable_set?

pl992 commented 4 years ago

indeed, thank you very much. Just a personal opinion, I still think it's quite important information and it isn't that easy to find in the documentation. Anyway I'd like to congratulate with all the developers for such a powerful and complete tool!