ValueError: too many values to unpack

[hovo1990]

Trying python alchemical_analysis.py -d gromacs/data/3-methylindole-38steps -q xvg -p dhdl -u kJ -v
and I get

/usr/lib/python2.7/dist-packages/scipy/optimize/minpack.py:236: RuntimeWarning: xtol=0.000000 is too small, no further improvement in the approximate
solution is possible.
warnings.warn(msg, RuntimeWarning)
Final dimensionless free energies
f_k =
[ 0. 2.85435075 5.36683423 7.54310132 9.41062672
11.00680904 12.36821823 13.52539 14.50082038 15.30933876
15.95876113 13.96406986 13.58387343 13.99782737 14.35144787
14.74395871 15.13605675 15.48968283 15.76312185 15.86488891
15.89546715 15.88720731 15.82995611 15.71117856 15.51663687
15.2322169 14.84601322 14.35037166 13.74694486 13.05530028
12.31496947 11.57396348 10.87411229 10.24438184 9.70162777
8.90365663 8.48915993 8.43014626]
MBAR initialization complete.
Traceback (most recent call last):
File "alchemical_analysis.py", line 1103, in
Deltaf_ij, dDeltaf_ij = estimatewithMBAR(u_kln, N_k, P.relative_tolerance, regular_estimate=True)
File "alchemical_analysis.py", line 239, in estimatewithMBAR
(Deltaf_ij, dDeltaf_ij) = MBAR.getFreeEnergyDifferences(uncertainty_method='svd-ew')
ValueError: too many values to unpack

[AMecklenfeld]

Hi,

I get the same error when using the tool on a cluster - on a local desktop pc it works just fine.
Have you been able to solve this problem?

[Lnaden]

Which version(s) of MBAR are you running with this script?

I think it was 2.0 of mbar.py where the returns in getFreeEnergyDifferences() changed from just Deltaf_ij, dDeltaf_jj to Deltaf_ij, dDeltaf_ij, Theta_ij where Theta_ij is the co-variance matrix. By default, Theta_ij is returned as None unless requested with the keyword so that the length and order of the returns object is consistent.

[schlaicha]

Hi,

the problem with the return value of MBAR.getFreeEnergyDifferences() still persists and can indeed be fixed by changing the line accordingly,
(Deltaf_ij, dDeltaf_ij, Theta_ij) = MBAR.getFreeEnergyDifferences(uncertainty_method='svd-ew').

However, the covariance matrix is not used at all in alchemical_analysis.py am I right?

Thanks for feedback! Alex

[Lnaden]

Correct, alchemical_analysis.py does not use the Theta_ij matrix. The only script in this repository that fetches the covariance is the harmonic-oscillators.py script for computeMultipleExpectations.