]> https://gitweb.dealii.org/ - dealii.git/commitdiff
Use the 'Set' variant of the compressed sparsity pattern as it doesn't show the quadr...
authorWolfgang Bangerth <bangerth@math.tamu.edu>
Tue, 4 Mar 2008 23:32:09 +0000 (23:32 +0000)
committerWolfgang Bangerth <bangerth@math.tamu.edu>
Tue, 4 Mar 2008 23:32:09 +0000 (23:32 +0000)
git-svn-id: https://svn.dealii.org/trunk@15853 0785d39b-7218-0410-832d-ea1e28bc413d

deal.II/examples/step-22/step-22.cc

index a75ec38080c635ea3437661ec07e127349de7f6e..33adb65ef18e8cc64c0143bad957dc09c33cf69c 100644 (file)
@@ -581,37 +581,92 @@ void StokesProblem<dim>::setup_dofs ()
                                   // sparsity pattern for the system
                                   // matrix we will create. We could
                                   // do this in the same way as in
-                                  // step-20, though there is a major
-                                  // reason not to do so. In 3D, the
-                                  // function
-                                  // <code>DoFTools::max_couplings_between_dofs</code>
-                                  // yields a conservative, large
-                                  // number for the coupling between
-                                  // the individual dofs, so that the
-                                  // memory initially provided for
-                                  // the creation of the sparsity
-                                  // pattern of the matrix is far too
-                                  // much -- so much actually that
-                                  // the initial sparsity pattern
-                                  // won't even fit into the physical
-                                  // memory of most systems already
-                                  // for moderately-sized 3D
-                                  // problems, see also the
-                                  // discussion in step-18.  Instead,
-                                  // we use a temporary object of the
-                                  // class
-                                  // BlockCompressedSparsityPattern,
-                                  // which is a block version of the
-                                  // compressed sparsity patterns
-                                  // from step-11 and step-18. All
-                                  // this is done inside a new scope,
-                                  // which means that the memory of
+                                  // step-20, i.e. directly build an
+                                  // object of type SparsityPattern
+                                  // through
+                                  // DoFTools::make_sparsity_pattern. However,
+                                  // there is a major reason not to
+                                  // do so: In 3D, the function
+                                  // DoFTools::max_couplings_between_dofs
+                                  // yields a conservative but rather
+                                  // large number for the coupling
+                                  // between the individual dofs, so
+                                  // that the memory initially
+                                  // provided for the creation of the
+                                  // sparsity pattern of the matrix
+                                  // is far too much -- so much
+                                  // actually that the initial
+                                  // sparsity pattern won't even fit
+                                  // into the physical memory of most
+                                  // systems already for
+                                  // moderately-sized 3D problems,
+                                  // see also the discussion in
+                                  // step-18.  Instead, we first
+                                  // build a temporary object that
+                                  // uses a different data structure
+                                  // that doesn't require allocating
+                                  // more memory than necessary but
+                                  // isn't suitable for use as a
+                                  // basis of SparseMatrix or
+                                  // BlockSparseMatrix objects; in a
+                                  // second step we then copy this
+                                  // object into an object of
+                                  // BlockSparsityPattern. This is
+                                  // entirely analgous to what we
+                                  // already did in step-11 and
+                                  // step-18.
+                                  //
+                                  // There is one snag again here,
+                                  // though: just as in step-27, it
+                                  // turns out that using the
+                                  // CompressedSparsityPattern (or
+                                  // the block version
+                                  // BlockCompressedSparsityPattern
+                                  // we would use here) turns out to
+                                  // have a bottleneck that makes the
+                                  // algorithm to build the sparsity
+                                  // pattern be quadratic in the
+                                  // number of degrees of
+                                  // freedom. This doesn't become
+                                  // noticable until we get well into
+                                  // the range of several 100,000
+                                  // degrees of freedom, but
+                                  // eventually dominates the setup
+                                  // of the linear system when we get
+                                  // to more than a million degrees
+                                  // of freedom. This is due to the
+                                  // data structures used in the
+                                  // CompressedSparsityPattern class,
+                                  // nothing that can easily be
+                                  // changed. Fortunately, there is
+                                  // an easy solution, as already
+                                  // pointed out in step-27: the
+                                  // CompressedSetSparsityPattern
+                                  // class (and its block variant
+                                  // BlockCompressedSetSparsityPattern)
+                                  // has exactly the same interface,
+                                  // uses a different internal data
+                                  // structure, is slightly slower
+                                  // for smaller numbers of degrees
+                                  // of freedom (but there we don't
+                                  // care that much anyway) but is
+                                  // linear in the number of degrees
+                                  // of freedom and therefore much
+                                  // more efficient for large
+                                  // problems.
+                                  //
+                                  // Consequently, this is the class
+                                  // that we will use for our
+                                  // intermediate sparsity
+                                  // representation. All this is done
+                                  // inside a new scope, which means
+                                  // that the memory of
                                   // <code>csp</code> will be
                                   // released once the information
                                   // has been copied to
                                   // <code>sparsity_pattern</code>.
   {
-    BlockCompressedSparsityPattern csp (2,2);
+    BlockCompressedSetSparsityPattern csp (2,2);
 
     csp.block(0,0).reinit (n_u, n_u);
     csp.block(1,0).reinit (n_p, n_u);

In the beginning the Universe was created. This has made a lot of people very angry and has been widely regarded as a bad move.

Douglas Adams


Typeset in Trocchi and Trocchi Bold Sans Serif.