If the correct average is provided, compress() may still
fail if more than the average numebr of entries are added
to an early row (e.g. the first one)

Due to a confusion between the number of lines N and the number of
columns M, Matrix::transpose() gave wrong results for matrices with
N != M.

Due to a bug in Matrix::transpose() it will give wrong results for
non-quadratic matrices. Thus the new test case fails for the moment.

Remove output that printed the matrix for parallel runs for small grids n<20.

According to git blame the last time these constructors were changed was
in 2011.

Simplifies the memory management.

Reimplement this using Hybrid::forEach. Along the way this
also fixes the bug that operator<< ignores the passed stream
and uses std::cout instead.

The new implementation is based on Hybrid::forEach

Test with non-zero matrix and print results

Reimplement this using Hybrid::forEach. This also adds some
comments and uses proper variable names to make the nan-tracking
easier to understand.

Reimplement this using Hybrid::forEach.

Reimplement this using Hybrid::forEach.

Reimplement this using Hybrid::forEach.

These separate impelmentations could be replaced by a single one,
once we have IsNumber. Currently we could use std::is_arithmetic
instead, but that's not the topic of this change.

Reimplement this using Hybrid::forEach.

Reimplement this using Hybrid::forEach. Along the way this
also fixes the bug that operator<< ignores the passed stream
and uses std::cout instead.

The implementation in MultiTypeBlockMatrix would be even
simpler, if it would provide a static size() method, because
we could then use a simple Hybrid::forEach over its entries
instead of one over all indices.

Except for those parts that do still rely on the enable trick to
test a certain feature set.

* Use correct macro
* adjust changes for other types then double, too

Arne Morten Kvarving authored 8 years ago and Christoph Grüninger committed 8 years ago

API for solver routine changed. Since there is no versioning
in SuperLU we have to do a probe to detect the new library.

Here we try to solve A*x=b for x, with
```
      / 1  1 \       / 1 \
  A = |      |,  b = |   |,
      \ 1  1 /       \ 2 /
```
which has no solution.  In the CGSolver, the this leads to a NaN defect after
46 iterations.  In BiCGSTABSolver this leads to "abs(h) < EPSILON" after 1.5
iterations.

BiCGSTABSolver and RestartedGMResSolver also throw SolverAbort when they
detect "breakdown", checking for that is left as a todo.

Fixes dune-common#23

Some tests assume MPI being present, although they are not executed
in parallel. This information got lost in the transition to dune_add_test
and is restored with this patch. Thanks to @markus.blatt for noticing.

Document that `BiCGSTABSolver` and `RestartedGMResSolver` can throw
`SolverAbort` when they detect a breakdown.

The previous commit made fastamg compile, but broke compilation of
matrixredisttest for me.  This makes matrixredisttest compile again.

SolverAbort is derived from ISTLError, so any existing catch-clause should
still apply.

SolverAbort is derived from ISTLError, so any existing catch-clause should
still apply.

recalculateHierarchy() assumed that there is information avaiable
that is only there in parallel runs. This failed miserably.

This commit fixes this be implementing dummy versions for
redistributing matrices in  a sequential run and does not
query any parallel stuff any more

This closes #11.

This allows other code to access field_type and real_type for a MultiTypeBlockVector
using the FieldTraits class.  Caveat: in principle, individual entries of a
MultiTypeBlockVector could use different field_types.  The implementation always
returns the first one.