Bug #407
make_edi
Description
When I try to use make_edi to generate a .edi file it works with
eigenvector files that were generated using g_covar -fit but not when I
generate the eigenvector file using g_covar -nofit on an input trajectory that
has already been fit using trjconv -fit. I get a segmentation fault instead. I am attaching sample eigenvector files generated using g_covar -fit
"evec170-1-172fit.trr "
and using g_covar -nofit
"evec170-1-172nfit.trr"
and the reference structure that I use for the fitting, also used as input for -s
"fframe170-1-172calpha.pdb"
Thanks
Vijaya
History
#1 Updated by vijaya empty about 10 years ago
Created an attachment (id=448)
input eigenvector file generated with g_covar -fit
#2 Updated by vijaya empty about 10 years ago
Created an attachment (id=449)
eigenvector file generated with g_covar -nofit
#3 Updated by vijaya empty about 10 years ago
Created an attachment (id=450)
input structure file for -s
#4 Updated by Erik Lindahl almost 10 years ago
This appears to be a problem in g_covar; the non-fitted eigenvectors are written out as all zeros.
#5 Updated by Erik Lindahl almost 10 years ago
Sometimes it helps to to try the problem with the latest code in git first (my bad). This was fixed by Carsten Kutzner already in january, but 4.0.7 was probably released before that. I've tested that it works in the git head branch too now.