pyhsmm
pyhsmm copied to clipboard
I took a shot at adding support for left-right models in this PR: https://github.com/mattjj/pyhsmm/pull/87. For reference, see https://github.com/hmmlearn/hmmlearn/issues/47. The idea is to fix entries in the transition matrix (and possibly...
The modifications fix the zeros present in the transition matrix and initial state probabilities to ensure the model is a left-right H(S)MM.
This fix allowed me to build and install the package on Windows using Visual Studio's cl.exe compiler. Based on https://github.com/lovell/farmhash/issues/4.
I test my model with the following setting and it throws an runtimewarning every time. I tested the model with the same setting in py2. Sometimes it's safe but sometimes...
Thanks for sharing your amazing library. We will be using it in an article, and in addition to citing the Johnson 2013 article we would also like to cite the...
I've got following error during resampling. /// error message Traceback (most recent call last): File "hmm.py", line 150, in model.resample_model() File "/usr/lib64/python2.7/site-packages/pyhsmm-0.1.6-py2.7-linux-x86_64.egg/pyhsmm/models.py", line 459, in resample_model self.resample_parameters() File "/usr/lib64/python2.7/site-packages/pyhsmm-0.1.6-py2.7-linux-x86_64.egg/pyhsmm/models.py", line...
The concentration parameter resampling strategy is suboptimal and possibly inaccurate. Try replacing it with either slice sampling or discretization.
I just found pyhsmm last week. First a minor issue: It looks like future is not installed by default in anaconda's distribution of python, so that pip install pyhsmm fails,...
Tried to use hmm.plot() in the svi example, and got the following:  Despite showing different emissions, they are all one color because only the latest state seems to be...
The weak limit transitions class does not support expectations necessary for mean field. We should either support those expectations (probably doesn't even make sense for "weak limit" models) or we...