search

MobleyLab / alchemical-analysis

An open tool implementing some recommended practices for analyzing alchemical free energy calculations
MIT License
114 stars 58 forks source link

small ligand solvation energy calculation:error:cannot compute statistical inefficiency #83

Open sanasaeed opened 8 years ago

sanasaeed commented 8 years ago

i have done ligand simulation in water and now wantto calculate solvation free energy , i used 31 lambdas states, i got this error by using alchemical_analysis.py.

Traceback (most recent call last): File "/share/anaconda2/bin/alchemical_analysis.py", line 4, in import('pkg_resources').run_script('alchemical-analysis==1.0.2.dev0', 'alchemical_analysis.py') File "/share/anaconda2/lib/python2.7/site-packages/setuptools-19.6.2-py2.7.egg/pkg_resources/init.py", line 724, in run_script File "/share/anaconda2/lib/python2.7/site-packages/setuptools-19.6.2-py2.7.egg/pkg_resources/init.py", line 1649, in run_script File "/share/anaconda2/lib/python2.7/site-packages/alchemical_analysis-1.0.2.dev0-py2.7.egg/EGG-INFO/scripts/alchemical_analysis.py", line 1226, in main() File "/share/anaconda2/lib/python2.7/site-packages/alchemical_analysis-1.0.2.dev0-py2.7.egg/EGG-INFO/scripts/alchemical_analysis.py", line 1197, in main dhdl, N_k, u_kln = uncorrelate(sta=numpy.zeros(K, int), fin=nsnapshots, do_dhdl=True) File "/share/anaconda2/lib/python2.7/site-packages/alchemical_analysis-1.0.2.dev0-py2.7.egg/EGG-INFO/scripts/alchemical_analysis.py", line 181, in uncorrelate g[k] = pymbar.timeseries.statisticalInefficiency(dhdl_sum) File "build/bdist.linux-x86_64/egg/pymbar/timeseries.py", line 178, in statisticalInefficiency pymbar.utils.ParameterError: Sample covariance sigma_AB^2 = 0 -- cannot compute statistical inefficiency

davidlmobley commented 8 years ago

Should have been fixed by #78, I think. Please verify you're using a version you just installed (i.e. in the last 20 days), or download and install the latest directly from source and confirm this persists?

Normally (prior to the fix) this issue arises only when you have a free energy difference which is identically zero between some (or all) of your lambda values, so in some sense the problem is with your data and not as much with the code, though the code OUGHT to handle this properly.

If you are using the latest version, can you provide a link to a Dropbox/Google Drive/etc. file with your data in it so we can troubleshoot?

sanasaeed commented 8 years ago

no i am using the older one. i will install the new one and will try again. where can i fine the description about the modifications you made in the new version. i want to know the difference from the older version. Thanks for your help

nathanmlim commented 8 years ago

Hi Sanaa, since we don't have a version scheme implemented the changes would depend on when you first installed alchemical analysis. You can also view the commit log if you want to get into the details. In general, there have been some important fixes to primarily address this issue you are seeing and some issues with MBAR error totaling.

sanasaeed commented 8 years ago

hi Nathan Limthanks for replyi just wanted to see what are the changes you made. i'll see commit log

Sana Saeed Khan,Teaching-Research AssistantChemoinformatics LabGraduate Student, MS bioinfoDepartment of BioinformaticsSoongsil University, Seoul, South Korea. 

On Thursday, March 10, 2016 11:15 AM, Nathan Lim <notifications@github.com> wrote:

Hi Sanaa, since we don't have a version scheme implemented the changes would depend on when you first installed alchemical analysis. You can also view the commit log if you want to get into the details. In general, there have been some important fixes to primarily address this issue you are seeing and some issues with MBAR error totaling.— Reply to this email directly or view it on GitHub.

davidlmobley commented 8 years ago

You should be able to get a fairly complete log of changes looking at the list of closed pull requests here https://github.com/MobleyLab/alchemical-analysis/pulls?q=is%3Apr+is%3Aclosed

You can also of course review detailed code changes associated with each.

Generally speaking, all of the changes have been bug fixes/extensions. I don't think we've actually changed details of how anything works except how decorrelation of correlated samples works in a very specific case.

Thanks.

On Wed, Mar 9, 2016 at 5:57 PM, Sana Saeed notifications@github.com wrote:

no i am using the older one. i will install the new one and will try again. where can i fine the description about the modifications you made in the new version. i want to know the difference from the older version. Thanks for your help

— Reply to this email directly or view it on GitHub https://github.com/MobleyLab/alchemical-analysis/issues/83#issuecomment-194617186 .

David Mobley dmobley@gmail.com 949-385-2436