Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Issue with numexpr() on MBAR initialization; expects to find mkl acceleration and crashes when not found

See original GitHub issue

Howdy folks,

Upon initializing MBAR using:

# Initialize MBAR with Newton-Raphson if (n==0): # only print this information the first time print "" print "Initializing MBAR:" print "--K = number of Temperatures with data = %d" % (originalK) print "--L = number of total Temperatures = %d" % (K) print "--N = number of Energies per Temperature = %d" % (numpy.max(Nall_k))

    # Use Adaptive Method (Both Newton-Raphson and Self-Consistent, testing which is better)
    if (n==0):
            initial_f_k = None # start from zero
            initial_f_k = mbar.f_k # start from the previous final free energies to speed convergence

    mbar = pymbar.MBAR(u_kln, Nall_k, verbose=False, relative_tolerance=1e-12, initial_f_k=initial_f_k)`

^^(Sorry for how this block of code looks. I couldn’t get it to look right in github. I assure you it is structurally correct in the actual python script)

I get an issue with the numexpr module trying to find a means for mkl acceleration, which gave the following error stack upon failing to do so.

pythonpython: : symbol lookup errorsymbol lookup error: pythonpython: /home/bmanubay/anaconda2/lib/python2.7/site-packages/numexpr/../../../ : /home/bmanubay/anaconda2/lib/python2.7/site-packages/numexpr/../../../ symbol lookup errorsymbol lookup errorpython: undefined symbol: mkl_serv_getenv: : : undefined symbol: mkl_serv_getenv/home/bmanubay/anaconda2/lib/python2.7/site-packages/numexpr/../../../libmkl_vml_avx2.sopython/home/bmanubay/anaconda2/lib/python2.7/site-packages/numexpr/../../../libmkl_vml_avx2.sopythonsymbol lookup error: python: : : : undefined symbol: mkl_serv_getenv: symbol lookup errorundefined symbol: mkl_serv_getenv

(very odd traceback…)

From what @mrshirts and I could tell, the origin is traced back to where the logsumexp() function calls numexpr.evaluate() on multiple occasions. We checked that numexpr() was functional on my python installation, but we think MBAR might be trying to implement some mkl option automatically without testing for it first. Any thoughts?

Issue Analytics

  • State:closed
  • Created 7 years ago
  • Comments:6 (2 by maintainers)

github_iconTop GitHub Comments

Lnadencommented, Jul 1, 2016

Right. tries to load numexpr which is part of the MKL acceleration library, and has a fallback if it can’t find it. So the issue would lie at that level.

mrshirtscommented, Feb 8, 2017

The current plan is to revert some of the algorithms back to not need mlk-rt. I just got back from travelling, and this is one of the things to do in the next couple weeks. Thanks for the feedback!

Read more comments on GitHub >

github_iconTop Results From Across the Web

Workaround when IDL to python bridge is crashing due to ...
The python to IDL bridge may work - but could also crash in some cases. The issue is most probably due to the...
Read more >
numexpr - Read the Docs
NumExpr requires Python 2.6 or greater, and NumPy 1.7 or greater. ... process in order to know whether MKL has been detected or...
Read more >
Rare crashes on MKL - Intel Communities
I have implemented in C++ an algorithm in image processing using (among other things) fftw wrappers in MKL library (version 2018.3.210) I am...
Read more >
Accelerating Python with Hardware-Optimized Computing
In this guide, we will cover: How to choose a BLAS and LAPACK distribution; How to install the MKL or OpenBLAS distributions; Troubleshooting ......
Read more >
FATAL ERROR mkl_vml_avx2.dll - Esri Community
I'm new in arcgis pro, I want to train a model for deep learning with python on jupyter notebooks, when running epoch 7,...
Read more >

github_iconTop Related Medium Post

No results found

github_iconTop Related StackOverflow Question

No results found

github_iconTroubleshoot Live Code

Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start Free

github_iconTop Related Reddit Thread

No results found

github_iconTop Related Hackernoon Post

No results found

github_iconTop Related Tweet

No results found

github_iconTop Related Post

No results found

github_iconTop Related Hashnode Post

No results found