<div dir="ltr"><div><div><div><div><div><div><div>Hi list<br><br></div>I am currently in the process of installing QE on a medium size cluster, as a replacement for DACAPO. Everything seems to be working fine while compiling the source code, but when i try a really simple example like the test.py below, I get a IO Error "Broken pipe" from espreeso.py:<br>
<br></div>Example:<br></div>test.py <br><br>#!/usr/bin/env python<br><br>from ase import optimize<br>from ase import Atoms<br>from espresso import espresso<br><br>a=Atoms('H2',[[0,0,0],[0.9,0,0]],cell=(3,3,3))<br>
calc = espresso(pw=400,dw=4000,kpts=(1,1,1),nbands=-5,xc='BEEF')<br>a.set_calculator(calc)<br><br>qn = optimize.QuasiNewton(a,trajectory='relax.traj')<br>qn.run(fmax=0.1)<br><br><br><br></div>Error message:<br>
self.cinp, self.cout = os.popen2('cd '+self.scratch+' ; '+'pw.x -in pw.inp')<br>At line 380 of file move_ions.f90 (unit = 5, file = 'fort.5')<br>Fortran runtime error: End of file<br>BFGSLineSearch: 0 10:53:16 -35.468850 0.2491<br>
Traceback (most recent call last):<br> File "test.py", line 14, in <module><br> qn.run(fmax=0.1)<br> File "/usr/lib/python2.6/site-packages/ase/optimize/optimize.py", line 119, in run<br> self.step(f)<br>
File "/usr/lib/python2.6/site-packages/ase/optimize/bfgslinesearch.py", line 104, in step<br> c2=self.c2, stpmax=self.stpmax)<br> File "/usr/lib/python2.6/site-packages/ase/utils/linesearch.py", line 57, in _line_search<br>
fval = func(xk + stp * pk, *args)<br> File "/usr/lib/python2.6/site-packages/ase/optimize/bfgslinesearch.py", line 153, in func<br> return calc.get_potential_energy(self.atoms,force_consistent=True) / self.alpha<br>
File "/usr/lib/python2.6/site-packages/ase/calculators/general.py", line 23, in get_potential_energy<br> self.update(atoms)<br> File "/opt/espresso/espresso.py", line 756, in update<br> self.read(atoms)<br>
File "/opt/espresso/espresso.py", line 783, in read<br> self.cinp.flush()<br>IOError: [Errno 32] Broken pipe<br><br><br><br></div>I've been checking permissions and paths on the scratch dirs and to the pseudopotentials and so on, but i haven't found the answer yet. Can anybody point me in the right direction?<br>
<br></div>Thanks in advance<br><br></div>Jesper R. Selknęs - Computer scientist at MESHTechnologies<br></div>