Closed trendelkampschroer closed 8 years ago
Thank you very much. Can you please also run the PyEMMA unit tests when using this implementation, to make sure that there are no secondary effects?
For the latest version of PyEMMA devel I get two errors realted to estimate_hidden_markov_model and timescales_hmsm.
I also get an import error for the name _get_indices_phi.
But these things don't seem to be related to the sampler, is that correct?
Latest version of pyemma needs mdtraj-1.5 and bhmm-0.5.2, please upgrade.
mdtraj is already version 1.5.1 on my machine, when I do easy_install --user bhmm==0.5.2 I get the following error
Processing bhmm-0.5.2.tar.gz Writing /tmp/easy_install-nO7Fki/bhmm-0.5.2/setup.cfg Running bhmm-0.5.2/setup.py -q bdist_egg --dist-dir /tmp/easy_install-nO7Fki/bhmm-0.5.2/egg-dist-tmp-d4D4l2 error: Setup script exited with error: unknown file type '.pyx' (from './bhmm/hidden/impl_c/hidden.pyx')
please try pip install bhmm --user
upgrade setuptools first: pip install setuptools -U --user; pip install bhmm --user
The check if setuptools has a sufficient version seems not to work as expected.
I've got bhmm-0.5.2. successfully installed, and the HMM-related errors are gone now.
Strangely I still get
ImportError: cannot import name _get_indices_phi
with a mdtraj version of 1.5.1
update your pyemma git/install to at least 2.0.2
Then you don't have PyEMMA 2.0.2.
Am 13/11/15 um 16:30 schrieb Benjamin Trendelkamp-Schroer:
I've got bhmm-0.5.2. successfully installed, and the HMM-related errors are gone now.
Strangely I still get
ImportError: cannot import name _get_indices_phi
with a mdtraj version of 1.5.1
— Reply to this email directly or view it on GitHub https://github.com/markovmodel/msmtools/pull/62#issuecomment-156464103.
Prof. Dr. Frank Noe Head of Computational Molecular Biology group Freie Universitaet Berlin
Phone: (+49) (0)30 838 75354 Web: research.franknoe.de
I have PyEMMA latest git devel. I can successfully import pyemma in ipython, the error occurs only if I run nosetests/pyemma. Maybe we can have a look at it together next week.
I suppose nose imports pyemma from the wrong location then. If you invoke nose in your git working copy, it should use the right version.
I have all tests on PyEMMA latest devel running successfully. Can someone please merge this pull-request.
Thank you!
…eds only check undirected connectivity of count matrix. ii) is_positive checks >=1e-15 instead of >= 1e-8. iii) Use sgamma instead of gengamma to generate Gamma-distributed random variates, possibly avoiding underflow/overflow due to small/large scaling factor beta.