markovmodel / pyemma_tutorials

How to analyze molecular dynamics data with PyEMMA
Creative Commons Attribution 4.0 International
72 stars 35 forks source link

Issue on clustering after changing the pip installation and trying conda installation #199

Closed andresilvapimentel closed 2 years ago

andresilvapimentel commented 3 years ago

I have used pyemma with no issue for over a year installing it on google colab with: pip install wheel pip install pyemma

Suddenly, it stopped working last week. I do not know why. I found the following issue:

Building wheels for collected packages: pyemma Building wheel for pyemma (PEP 517) ... error ERROR: Failed building wheel for pyemma Failed to build pyemma ERROR: Could not build wheels for pyemma which use PEP 517 and cannot be installed directly

Then, I tried to install it by using the conda package on google colab bu using the following:

! wget https://repo.anaconda.com/miniconda/Miniconda3-py37_4.8.2-Linux-x86_64.sh ! chmod +x Miniconda3-py37_4.8.2-Linux-x86_64.sh ! bash ./Miniconda3-py37_4.8.2-Linux-x86_64.sh -b -f -p /usr/local import sys sys.path.append('/usr/local/lib/python3.7/site-packages/')

!curl -s https://raw.githubusercontent.com/markovmodel/PyEMMA/devel/install_miniconda%2Bpyemma.sh | bash !conda install -c conda-forge pyemma

It works pretty well until clustering. I found the following using: cluster = pyemma.coordinates.cluster_kmeans(tica_output, k=k, max_iter=50, stride=50)

/usr/local/lib/python3.7/site-packages/deeptime/data/_pbf_simulator.py in () 3 from threadpoolctl import threadpool_limits 4 ----> 5 from . import _data_bindings as bd 6 from ..util.decorators import plotting_function 7 from ..util.parallel import handle_n_jobs, joining

ImportError: /usr/lib/x86_64-linux-gnu/libstdc++.so.6: version `GLIBCXX_3.4.26' not found (required by /usr/local/lib/python3.7/site-packages/deeptime/data/_data_bindings.cpython-37m-x86_64-linux-gnu.so)

Please, let me know what is going on. hanks.

clonker commented 3 years ago

We probably have to do a patch-release of pyemma. In the meantime you can try this: !pip install git+https://github.com/markovmodel/pyemma.git@devel

clonker commented 3 years ago

patch is out, would you mind trying if everything works for you now? the pyemma version you are looking for is 2.5.10.