<div dir="ltr"><div dir="ltr">Is this relly necessary, because in the last sentences of the chapter it states that:</div><div dir="ltr">An additional benefit of multi-communicator support is that it enables parallel spectrum slicing runs without the need to install a parallel direct solver (MUMPS). The following commandline example uses sequential linear solves in 4 partitions, one process each:<div>Therefore i assumed that it is not necessary to compile PETsc4py with an external solver e.g. MUMPS</div></div></div><br><div class="gmail_quote"><div dir="ltr">Am Mo., 17. Sep. 2018 um 10:47 Uhr schrieb Jose E. Roman <<a href="mailto:jroman@dsic.upv.es">jroman@dsic.upv.es</a>>:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">You need a parallel direct solver such as MUMPS. This is explained in section 3.4.5.<br>
Jose<br>
<br>
<br>
> El 17 sept 2018, a las 10:41, Jan Grießer <<a href="mailto:griesser.jan@googlemail.com" target="_blank">griesser.jan@googlemail.com</a>> escribió:<br>
> <br>
> def solve_eigensystem(DynMatrix_nn, Unity_nn, Dimension, LowerLimit, UpperLimit):<br>
> # Create the EPS solver <br>
> E = SLEPc.EPS().create()<br>
> <br>
> # Create the preconditioner and set it to Cholesky<br>
> pc = PETSc.PC().create()<br>
> pc.setType(pc.Type.CHOLESKY)<br>
> <br>
> # Create the KSP object<br>
> ksp = PETSc.KSP().create()<br>
> ksp.setType(ksp.Type.PREONLY)<br>
> ksp.setPC(pc)<br>
> <br>
> # Set up the spectral transformations <br>
> st = SLEPc.ST().create()<br>
> st.setType("sinvert")<br>
> st.setKSP(ksp)<br>
> # Setup spectral transformation <br>
> E.setST(st)<br>
> <br>
> # Eigenvalues should be real, therefore we start to order them from the smallest real value |l.real|<br>
> E.setWhichEigenpairs(E.Which.ALL)<br>
> # Set the interval of spectrum slicing<br>
> E.setInterval(LowerLimit, UpperLimit)<br>
> # Since the dynamical matrix is symmetric and real it is hermitian. Use GHEP for the spectrum slicing. Operatormatrix B is just a unit matrix <br>
> E.setProblemType(SLEPc.EPS.ProblemType.GHEP)<br>
> # Use the Krylov Schur method to solve the eigenvalue problem<br>
> E.setType(E.Type.KRYLOVSCHUR)<br>
> # Partition the Krylov schnur problem in npart procceses<br>
> E.setKrylovSchurPartitions(10)<br>
> # Set the convergence criterion to relative to the eigenvalue and the maximal number of iterations<br>
> E.setConvergenceTest(E.Conv.REL)<br>
> E.setTolerances(tol = 1e-7, max_it = 1000)<br>
> # Set the matrix in order to solve <br>
> E.setOperators(DynMatrix_nn, Unity_nn)<br>
> # Sets EPS options from the options database. <br>
> E.setFromOptions()<br>
> # Sets up all the internal data structures necessary for the execution of the eigensolver.<br>
> E.setUp()<br>
> <br>
> # Solve eigenvalue problem <br>
> startClock = time.clock()<br>
> startTime = time.time()<br>
> E.solve()<br>
> <br>
> Has maybe one of you any idea why this happens and where the problem is ? <br>
> <br>
> Am Mo., 17. Sep. 2018 um 10:40 Uhr schrieb Jan Grießer <<a href="mailto:griesser.jan@googlemail.com" target="_blank">griesser.jan@googlemail.com</a>>:<br>
> I am aware that SLEPc is not supposed to calculate all eigenvalues and eigenvectors, my problem is simply that i want for a physical large enough system all of them before i can make the transition to go to the smallest ones. <br>
> Competitiveness is of secondary importance at the moment. <br>
> But ihave a problem connected with spectrum slicing. I followed the instructions in the manual of Chap. 3.4.5 Spectrum Slicing and converted them to the python package. <br>
> But now i get the following error. It appears to me that it is not able to find the ksp object, but i actually do not know why this is the case. <br>
> aceback (most recent call last):<br>
> File "Eigensolver_spectrum_slicing.py", line 216, in <module><br>
> solve_eigensystem(DynMatrix_nn, Unity_nn, D_nn.shape, opt_dict.LowLimit, opt_dict.UpperLimit)<br>
> File "Eigensolver_spectrum_slicing.py", line 121, in solve_eigensystem<br>
> E.setUp()<br>
> File "SLEPc/EPS.pyx", line 1099, in slepc4py.SLEPc.EPS.setUp<br>
> petsc4py.PETSc.Error: error code 92<br>
> [14] EPSSetUp() line 165 in /tmp/pip-install-golhudw7/slepc/src/eps/interface/epssetup.c<br>
> [14] EPSSetUp_KrylovSchur() line 146 in /tmp/pip-install-golhudw7/slepc/src/eps/impls/krylov/krylovschur/krylovschur.c<br>
> [14] EPSSetUp_KrylovSchur_Slice() line 410 in /tmp/pip-install-golhudw7/slepc/src/eps/impls/krylov/krylovschur/ks-slice.c<br>
> [14] EPSSliceGetEPS() line 300 in /tmp/pip-install-golhudw7/slepc/src/eps/impls/krylov/krylovschur/ks-slice.c<br>
> [14] EPSSetUp() line 165 in /tmp/pip-install-golhudw7/slepc/src/eps/interface/epssetup.c<br>
> [14] EPSSetUp_KrylovSchur() line 146 in /tmp/pip-install-golhudw7/slepc/src/eps/impls/krylov/krylovschur/krylovschur.c<br>
> [14] EPSSetUp_KrylovSchur_Slice() line 461 in /tmp/pip-install-golhudw7/slepc/src/eps/impls/krylov/krylovschur/ks-slice.c<br>
> [14] EPSSliceGetInertia() line 331 in /tmp/pip-install-golhudw7/slepc/src/eps/impls/krylov/krylovschur/ks-slice.c<br>
> [14] STSetUp() line 271 in /tmp/pip-install-golhudw7/slepc/src/sys/classes/st/interface/stsolve.c<br>
> [14] STSetUp_Sinvert() line 132 in /tmp/pip-install-golhudw7/slepc/src/sys/classes/st/impls/sinvert/sinvert.c<br>
> [14] KSPSetUp() line 381 in /tmp/pip-install-xmiaat2t/petsc/src/ksp/ksp/interface/itfunc.c<br>
> [14] PCSetUp() line 923 in /tmp/pip-install-xmiaat2t/petsc/src/ksp/pc/interface/precon.c<br>
> [14] PCSetUp_Cholesky() line 86 in /tmp/pip-install-xmiaat2t/petsc/src/ksp/pc/impls/factor/cholesky/cholesky.c<br>
> [14] MatGetFactor() line 4318 in /tmp/pip-install-xmiaat2t/petsc/src/mat/interface/matrix.c<br>
> [14] See <a href="http://www.mcs.anl.gov/petsc/documentation/linearsolvertable.html" rel="noreferrer" target="_blank">http://www.mcs.anl.gov/petsc/documentation/linearsolvertable.html</a> for possible LU and Cholesky solvers<br>
> [14] Could not locate a solver package. Perhaps you must ./configure with --download-<package><br>
> <br>
> The code i used to solve the problem is <br>
> <br>
> Am Fr., 14. Sep. 2018 um 18:34 Uhr schrieb Matthew Knepley <<a href="mailto:knepley@gmail.com" target="_blank">knepley@gmail.com</a>>:<br>
> On Fri, Sep 14, 2018 at 12:19 PM Jose E. Roman <<a href="mailto:jroman@dsic.upv.es" target="_blank">jroman@dsic.upv.es</a>> wrote:<br>
> El 14 sept 2018, a las 17:45, Jan Grießer <<a href="mailto:griesser.jan@googlemail.com" target="_blank">griesser.jan@googlemail.com</a>> escribió:<br>
> <br>
>> Hey there,<br>
>> first i want to say thanks to Satish and Matt for helping with with my last problem with the mpi compilation. I have two questions related to solving a big, hermitian, standard eigenvalue problem using SLEPc4py., compiled with Intel MKL and Intel MPI. <br>
>> <br>
>> - I am using slepc4py with <br>
>> mpi and run it with around -n 20 cores at the moment and how i wanted to ask if there is an easy way to retrieve the eigenvectors? When i run my code and print for i in range(nconv):<br>
>> for i in range(nconv):<br>
>> <br>
>> val = E.<br>
>> getEigenpair(i, vr<br>
>> , vi)<br>
>> Print(<br>
>> vr.getArray())<br>
>> i get the parts of the eigenvectors according to the partition of the matrix. Is there any easy way to put them together in an array and write them to file ? (I am struggling a little bit with the building them in the correct order)<br>
> <br>
> You need VecScatterCreateToZero. There must be an equivalent in python.<br>
> <br>
> An alternative to this which you should consider, because it is simpler, is to write the vector to a file<br>
> using some format that PETSc understands, Then you just need vr.view(viewer) for a viewer like<br>
> the binary viewer or some ASCII format you like.<br>
> <br>
> Thanks,<br>
> <br>
> Matt<br>
>> - I need to solve eigenvalue problems up to a dimension of 100000 degrees of freedom and i need all eigenvalues and eigenvectors. I think solving all eigenvalues in one process is far too much and i thought about if it is possible to apply the spectrum slicing described in Chap. 3.4.5. Due to the nature of my problem, i am able to simulate smaller systems of 10000 DOF and extract the biggest eigenvalue, which will be the same for larger systems sizes. Is this in general possible since i have a standard HEP problem or is there a better and faster possibility to do this?<br>
> <br>
> In general, SLEPc is not intended for computing the whole spectrum. You can try with spectrum slicing but this will be competitive if computing just a percentage of eigenvalues, 50% say. <br>
> <br>
> Jose<br>
> <br>
>> <br>
>> Thank you very much!<br>
> <br>
> <br>
> -- <br>
> What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>
> -- Norbert Wiener<br>
> <br>
> <a href="https://www.cse.buffalo.edu/~knepley/" rel="noreferrer" target="_blank">https://www.cse.buffalo.edu/~knepley/</a><br>
<br>
</blockquote></div>