Différences

Ci-dessous, les différences entre deux révisions de la page.

Lien vers cette vue comparative

Les deux révisions précédentes Révision précédente
Prochaine révision
Révision précédente
fenics:2017 [2018/06/27 14:42] – adminfenics:2017 [2018/06/28 09:29] (Version actuelle) – [Batch Mode with SGE] admin
Ligne 3: Ligne 3:
 The new Fenics version is installed within Singularity container.  The new Fenics version is installed within Singularity container. 
 ===== How to use Fenics (interactive mode) ===== ===== How to use Fenics (interactive mode) =====
 +<note tip>We use ''mesoshared.univ-fcomte.fr'' node for interactive computation.</note> 
  
 First load the singularity container: First load the singularity container:
Ligne 18: Ligne 19:
 Or simply run: Or simply run:
 <code  bash> <code  bash>
-$ $SING_SPOOL/fenics.img python poisson.py+$ $SING_SPOOL/fenics.img poisson.py
 </code> </code>
  
-<note warning>Automatic plotting is disabled. At the end of the execution A la use  paraview (''module load tools/paraview'') to  plot results.+<note important> 
 +Since Fenics is installed for sequential usage. Please ignore the following warning message. 
 + 
 +''[[62548,1],0]: A high-performance Open MPI point-to-point messaging module 
 +was unable to find any relevant network interfaces: 
 + 
 +Module: OpenFabrics (openib) 
 +  Host: mesoshared 
 + 
 +Another transport will be used instead, although this may result in 
 +lower performance. 
 +''
  
 </note> </note>
 +
 +
  
 =====Batch Mode with SGE ===== =====Batch Mode with SGE =====
-Example of sGE SGE script to use.+Example of SGE script to use.
 <code bash> <code bash>
 #!/bin/bash #!/bin/bash
Ligne 37: Ligne 51:
 module load tools/singularity  module load tools/singularity 
    
-export PATH=$WORK+export HOME=$WORK
    
 singularity exec $SING_SPOOL/fenics.img python poisson.py singularity exec $SING_SPOOL/fenics.img python poisson.py
Ligne 45: Ligne 59:
 $ qsub script.sge $ qsub script.sge
 </code> </code>
 +
 +===== Program Example (poisson.py) =====
 +
 +<file  python poisson.py>
 +"""
 +FEniCS tutorial demo program: Poisson equation with Dirichlet conditions.
 +Test problem is chosen to give an exact solution at all nodes of the mesh.
 +
 +  -Laplace(u) = f    in the unit square
 +            u = u_D  on the boundary
 +
 +  u_D = 1 + x^2 + 2y^2
 +    f = -6
 +"""
 +
 +from __future__ import print_function
 +from fenics import *
 +
 +# Create mesh and define function space
 +mesh = UnitSquareMesh(8, 8)
 +V = FunctionSpace(mesh, 'P', 1)
 +
 +# Define boundary condition
 +u_D = Expression('1 + x[0]*x[0] + 2*x[1]*x[1]', degree=2)
 +
 +def boundary(x, on_boundary):
 +    return on_boundary
 +
 +bc = DirichletBC(V, u_D, boundary)
 +
 +# Define variational problem
 +u = TrialFunction(V)
 +v = TestFunction(V)
 +f = Constant(-6.0)
 +a = dot(grad(u), grad(v))*dx
 +L = f*v*dx
 +
 +# Compute solution
 +u = Function(V)
 +solve(a == L, u, bc)
 +
 +# Plot solution and mesh
 +plot(u)
 +plot(mesh)
 +
 +# Save solution to file in VTK format
 +vtkfile = File('poisson/solution.pvd')
 +vtkfile << u
 +
 +# Compute error in L2 norm
 +error_L2 = errornorm(u_D, u, 'L2')
 +
 +# Compute maximum error at vertices
 +vertex_values_u_D = u_D.compute_vertex_values(mesh)
 +vertex_values_u = u.compute_vertex_values(mesh)
 +import numpy as np
 +error_max = np.max(np.abs(vertex_values_u_D - vertex_values_u))
 +
 +# Print errors
 +print('error_L2  =', error_L2)
 +print('error_max =', error_max)
 +
 +# Hold plot
 +# interactive() //OLD methode to plot
 +
 +import matplotlib.pyplot as plt
 +plot(u)
 +plt.show()
 +
 +</file>
  • fenics/2017.1530110552.txt.gz
  • Dernière modification : 2018/06/27 14:42
  • de admin