Re: [AMBER] Can anyone point out whether my ambertools and amber test failures are significant

From: Jason Swails <>
Date: Sun, 01 Apr 2012 11:54:05 -0400

A couple notes here:

First, several PBSA tests are expected to fail in the Amber test suite.
This is due to a change between variable names in AmberTools 1.4 (which
was released with Amber 11) and AmberTools 1.5 (which was released a
year later). A full list of the failing tests can be found by running
the script in $AMBERHOME (alternatively, you can just
look at the script itself at the bottom).

As for the tests themselves, you can post them here, but I suggest
analyzing them yourself first. Round-off errors usually manifest
themselves in the last 1 or 2 decimal places reported. If the diffs are
greater than this, there is likely an error with that particular
functionality. Errors are more likely on Cygwin than Linux or Mac OS X
because it's tested less frequently and is really just a Unix emulator
for Windows with several limitations (e.g. a severely limited stack size
for one).

You should check the test failures and errors to make sure that the
functionality you're trying to use is properly working.


On Sun, 2012-04-01 at 04:17 -0700, Acoot Brett wrote:
> Dear All,
> I just completed the amber 11 and ambertools 1.5 installation (in cygwin) and I finished the test.
> The summary of the ambertools test:
> 384 file comparisons passed
> 19 file comparisons failed
> 7 tests experienced errors
> Test log file saved as logs/test_at_serial/2012-04-01_19-57-50.log
> Test diffs file saved as logs/test_at_serial/2012-04-01_19-57-50.diff
> The summary of the amber test:
> 359 file comparisons passed
> 16 file comparisons failed
> 7 tests experienced errors
> Test log file saved as logs/test_amber_serial/2012-04-01_20-37-37.log
> Test diffs file saved as logs/test_amber_serial/2012-04-01_20-37-37.diff
> Currently I do not have the ability to make the conclusions on that whether the failaures are due to round errors or some other reasons, and whether the failures are significant which can influence my normal function of ambertools and amber.
> Thus I want to e-mail my log files and diffs files to experts or users of AMBER to make the judgment.
> Will you please send me an e-mail to et me know you can make this Judgment? And then I will forward the files to you for your feedback.
> I am looking forward to getting a reply from you.
> Cheers,
> Acoot
> _______________________________________________
> AMBER mailing list

Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
AMBER mailing list
Received on Sun Apr 01 2012 - 09:00:04 PDT
Custom Search