Fipy

------------------------

The bump in major version number reflects more on the substantial increase
in capabilities and ease of use than it does on a break in compatibility
with FiPy 2.x. Few, if any, changes to your existing scripts should be
necessary.

The significant changes since version 2.1 are:

- :ref:`CoupledEquations` are now supported.
- A more robust mechanism for specifying :ref:`BoundaryConditions` is now
used.
- Most :class:`~fipy.meshes.mesh.Mesh`\es can be partitioned by
:ref:`MeshingWithGmsh`.
- :ref:`PYAMG` and :ref:`SCIPY` have been added to the :ref:`SOLVERS`.
- FiPy is capable of running under :term:`Python 3`.
- "getter" and "setter" methods have been pervasively changed to Python
properties.
- The test suite now runs much faster.
- Tests can now be run on a full install using `fipy.test()`.
- The functions of the :mod:`~fipy.tools.numerix` module are no longer
included in the :mod:`fipy` namespace. See
:mod:`examples.updating.update2_0to3_0` for details.
- Equations containing a :class:`~fipy.terms.transientTerm.TransientTerm`,
must specify the timestep by passing a ``dt=`` argument when calling
:meth:`~fipy.terms.term.Term.solve` or :meth:`~fipy.terms.term.Term.sweep`.

.. warning::

:term:`FiPy` 3 brought unavoidable syntax changes from :term:`FiPy` 2.
Please see :mod:`examples.updating.update2_0to3_0` for guidance on the
changes that you will need to make to your :term:`FiPy` 2.x scripts.

Fixes
-----

- `332 <https://github.com/usnistgov/fipy/issues/332>`_:
Inline failure on Ubuntu x86_64
- `324 <https://github.com/usnistgov/fipy/issues/324>`_:
constraining values with `ImplicitSourceTerm` not documented?
- `317 <https://github.com/usnistgov/fipy/issues/317>`_:
`gmshImport` tests fail on Windows due to shared file
- `316 <https://github.com/usnistgov/fipy/issues/316>`_:
changes to `gmshImport.py` caused `--inline` problems
- `313 <https://github.com/usnistgov/fipy/issues/313>`_:
Gmsh I/O
- `307 <https://github.com/usnistgov/fipy/issues/307>`_:
Failures on sandbox under buildbot
- `306 <https://github.com/usnistgov/fipy/issues/306>`_:
Add in parallel buildbot testing on more than 2 processors
- `302 <https://github.com/usnistgov/fipy/issues/302>`_:
`CellVariable.min()` broken in parallel
- `301 <https://github.com/usnistgov/fipy/issues/301>`_:
`Epetra.PyComm()` broken on Debian
- `300 <https://github.com/usnistgov/fipy/issues/300>`_:
`examples/cahnHilliard/mesh2D.py` broken with -- trilinos
- `299 <https://github.com/usnistgov/fipy/issues/299>`_:
Viewers not working when plotting meshes with zero cells in parallel
- `298 <https://github.com/usnistgov/fipy/issues/298>`_:
Memory consumption growth with repeated meshing, especially with
Gmsh
- `294 <https://github.com/usnistgov/fipy/issues/294>`_:
`--pysparse --inline` failures
- `293 <https://github.com/usnistgov/fipy/issues/293>`_:
`python examples/cahnHilliard/sphere.py --inline` segfaults on OS X
- `292 <https://github.com/usnistgov/fipy/issues/292>`_:
two `--scipy` failures
- `290 <https://github.com/usnistgov/fipy/issues/290>`_:
Improve test reporting to avoid inconsequential buildbot failures
- `288 <https://github.com/usnistgov/fipy/issues/288>`_:
gmsh importer and gmsh tests don't clean up after themselves
- `286 <https://github.com/usnistgov/fipy/issues/286>`_:
get running in Py3k
- `285 <https://github.com/usnistgov/fipy/issues/285>`_:
remove deprecated `viewers.make()`
- `284 <https://github.com/usnistgov/fipy/issues/284>`_:
remove deprecated `Variable.transpose()`
- `281 <https://github.com/usnistgov/fipy/issues/281>`_:
remove deprecated `NthOrderDiffusionTerm`
- `280 <https://github.com/usnistgov/fipy/issues/280>`_:
remove deprecated `diffusionTerm=` argument to `ConvectionTerm`
- `277 <https://github.com/usnistgov/fipy/issues/277>`_:
remove deprecated `steps=` from Solver
- `273 <https://github.com/usnistgov/fipy/issues/273>`_:
Make `DiffusionTermNoCorrection` the default
- `270 <https://github.com/usnistgov/fipy/issues/270>`_:
tests take *too* long!!!
- `267 <https://github.com/usnistgov/fipy/issues/267>`_:
Reduce the run times for chemotaxis tests
- `264 <https://github.com/usnistgov/fipy/issues/264>`_:
HANG in parallel test of `examples/chemotaxis/input2D.py` on some
configurations
- `261 <https://github.com/usnistgov/fipy/issues/261>`_:
`GmshImport` should read element colors
- `260 <https://github.com/usnistgov/fipy/issues/260>`_:
`GmshImport` should support all element types
- `259 <https://github.com/usnistgov/fipy/issues/259>`_:
Introduce `mesh.x` as shorthand for `mesh.cellCenters[0]` etc
- `258 <https://github.com/usnistgov/fipy/issues/258>`_:
`GmshExport` is not tested and does not work
- `252 <https://github.com/usnistgov/fipy/issues/252>`_:
Include Benny's improved interpolation patch
- `250 <https://github.com/usnistgov/fipy/issues/250>`_:
TeX is wrong in `examples.phase.quaternary`
- `247 <https://github.com/usnistgov/fipy/issues/247>`_:
`diffusionTerm(var=var1).solver(var=var0)` should fail sensibly
- `243 <https://github.com/usnistgov/fipy/issues/243>`_:
close out reconstrain branch
- `242 <https://github.com/usnistgov/fipy/issues/242>`_:
update documentation
- `240 <https://github.com/usnistgov/fipy/issues/240>`_:
Profile and merge reconstrain branch
- `237 <https://github.com/usnistgov/fipy/issues/237>`_:
`--Trilinos --no-pysparse` uses Pysparse?!?
- `236 <https://github.com/usnistgov/fipy/issues/236>`_:
anisotropic diffusion and constraints don't mix
- `235 <https://github.com/usnistgov/fipy/issues/235>`_:
changed constraints don't propagate
- `231 <https://github.com/usnistgov/fipy/issues/231>`_:
`factoryMeshes.py` not up to date with respect to keyword arguments
- `223 <https://github.com/usnistgov/fipy/issues/223>`_:
mesh in FiPy name space
- `218 <https://github.com/usnistgov/fipy/issues/218>`_:
Absence of `enthought.tvtk` causes test failures
- `216 <https://github.com/usnistgov/fipy/issues/216>`_:
Fresh FiPy gives "`ImportError: No viewers found`"
- `213 <https://github.com/usnistgov/fipy/issues/213>`_:
PyPI is failing
- `206 <https://github.com/usnistgov/fipy/issues/206>`_:
`gnuplot1d` gives error on plot of `FaceVariable`
- `205 <https://github.com/usnistgov/fipy/issues/205>`_:
wrong cell to cell normal in periodic meshes
- `203 <https://github.com/usnistgov/fipy/issues/203>`_:
Give helpful error on - or / of meshes
- `202 <https://github.com/usnistgov/fipy/issues/202>`_:
mesh manipulation of periodic meshes leads to errors
- `201 <https://github.com/usnistgov/fipy/issues/201>`_:
Use physical velocity in the manual/FAQ
- `200 <https://github.com/usnistgov/fipy/issues/200>`_:
FAQ gives bad guidance for anisotropic diffusion
- `195 <https://github.com/usnistgov/fipy/issues/195>`_:
term multiplication changes result
- `163 <https://github.com/usnistgov/fipy/issues/163>`_:
Default time steps should be infinite
- `162 <https://github.com/usnistgov/fipy/issues/162>`_:
remove ones and zeros from `numerix.py`
- `130 <https://github.com/usnistgov/fipy/issues/130>`_:
tests should be run with `fipy.tests()`
- `86 <https://github.com/usnistgov/fipy/issues/86>`_:
Grids should take `Lx`, `Ly`, `Lz` arguments
- `77 <https://github.com/usnistgov/fipy/issues/77>`_:
`CellVariable.hasOld()` should set `self.old`
- `44 <https://github.com/usnistgov/fipy/issues/44>`_:
Navier-Stokes

--------------------------

--------------------------

Fixes
-----

- `282 <https://github.com/usnistgov/fipy/issues/282>`_:
remove deprecated getters and setters
- `279 <https://github.com/usnistgov/fipy/issues/279>`_:
remove deprecated `fipy.meshes.numMesh` submodule
- `278 <https://github.com/usnistgov/fipy/issues/278>`_:
remove deprecated forms of Gmsh meshes
- `268 <https://github.com/usnistgov/fipy/issues/268>`_:
Set up `Zizou` as a working slave
- `262 <https://github.com/usnistgov/fipy/issues/262>`_:
issue with solvers
- `256 <https://github.com/usnistgov/fipy/issues/256>`_:
`Grid1D(dx=(1,2,3))` failure
- `251 <https://github.com/usnistgov/fipy/issues/251>`_:
parallel is broken
- `241 <https://github.com/usnistgov/fipy/issues/241>`_:
Set Sandbox up as a working slave
- `238 <https://github.com/usnistgov/fipy/issues/238>`_:
`_BinaryTerm.var` is not predictable
- `233 <https://github.com/usnistgov/fipy/issues/233>`_:
coupled convection-diffusion always treated as Upwind
- `224 <https://github.com/usnistgov/fipy/issues/224>`_:
"matrices are not aligned" errors in example test suite
- `222 <https://github.com/usnistgov/fipy/issues/222>`_:
Non-uniform `Grid3D` fails to __add__
- `221 <https://github.com/usnistgov/fipy/issues/221>`_:
Problem with fipy and gmsh
- `219 <https://github.com/usnistgov/fipy/issues/219>`_:
matforge css is hammer-headed
- `208 <https://github.com/usnistgov/fipy/issues/208>`_:
numpy 2.0: `arrays have a dot method`
- `207 <https://github.com/usnistgov/fipy/issues/207>`_:
numpy 2.0: `masked arrays cast right of product to ndarray`
- `196 <https://github.com/usnistgov/fipy/issues/196>`_:
Pysparse won't import in Python 2.6.5 on Windows
- `152 <https://github.com/usnistgov/fipy/issues/152>`_:
(Re)Implement SciPy solvers
- `138 <https://github.com/usnistgov/fipy/issues/138>`_:
FAQ on boundary conditions
- `100 <https://github.com/usnistgov/fipy/issues/100>`_:
testing from the Windows dist using the ipython command line
- `80 <https://github.com/usnistgov/fipy/issues/80>`_:
Windows - testing - idle `-ipython`
- `46 <https://github.com/usnistgov/fipy/issues/46>`_:
Variable needs to consider boundary conditions
- `45 <https://github.com/usnistgov/fipy/issues/45>`_:
Slicing a vector Variable should produce a scalar Variable

--------------------------

--------------------------

The significant changes since version 2.1.1 are:

- :term:`Trilinos` efficiency improvements
- Diagnostics of the parallel environment

Fixes
-----

- `232 <https://github.com/usnistgov/fipy/issues/232>`_:
Mayavi broken on windows because it has no `SIGHUP`.
- `230 <https://github.com/usnistgov/fipy/issues/230>`_:
`factoryMeshes.py` not up to date with respect to keyword arguments
- `226 <https://github.com/usnistgov/fipy/issues/226>`_:
`MatplotlibViewer` fails if backend doesn't support `flush_events()`
- `225 <https://github.com/usnistgov/fipy/issues/225>`_:
Windows interactive plotting mostly broken
- `217 <https://github.com/usnistgov/fipy/issues/217>`_:
Gmsh `CellVariables` can't be unpickled
- `191 <https://github.com/usnistgov/fipy/issues/191>`_:
`sphereDaemon.py` missing in FiPy 2.1 and from trunk
- `187 <https://github.com/usnistgov/fipy/issues/187>`_:
Concatenated `Mesh` garbled by `dump.write`/`read`

--------------------------

--------------------------

The significant changes since version 2.1 are:

- :class:`~fipy.viewers.matplotlibViewer.MatplotlibViewer` can display
into an existing set of Matplotlib axes.

- :term:`Pysparse` and :term:`Trilinos` are now completely independent.

Fixes
-----

- `199 <https://github.com/usnistgov/fipy/issues/199>`_:
dummy viewer results in
"`NotImplementedError: can't instantiate abstract base class`"
- `198 <https://github.com/usnistgov/fipy/issues/198>`_:
bug problem with `CylindricalGrid1D`
- `197 <https://github.com/usnistgov/fipy/issues/197>`_:
How to tell if parallel is configured properly?
- `194 <https://github.com/usnistgov/fipy/issues/194>`_:
`FIPY_DISPLAY_MATRIX` on empty matrix with large b-vector throws
`ValueError`
- `193 <https://github.com/usnistgov/fipy/issues/193>`_:
`FIPY_DISPLAY_MATRIX` raises `ImportError` in FiPy 2.1 and trunk
- `192 <https://github.com/usnistgov/fipy/issues/192>`_:
`FIPY_DISPLAY_MATRIX=terms` raises `TypeError` in FiPy 2.1 and trunk

------------------------

------------------------

The relatively small change in version number belies significant advances
in :term:`FiPy` capabilities. This release did not receive a "full"
version increment because it is completely (er... [almost]_) compatible
with older scripts.

The significant changes since version 2.0.2 are:

- :term:`FiPy` can use :term:`Trilinos` for :ref:`PARALLEL`.

- We have switched from :term:`MayaVi` 1 to :term:`Mayavi` 2. This
:class:`~fipy.viewers.viewer.Viewer` is an independent process that
allows interaction with the display while a simulation is running.

- Documentation has been switched to :term:`Sphinx`, allowing the entire manual
to be available on the web and for our documentation to link to the
documentation for packages such as :mod:`numpy`, :mod:`scipy`,
:mod:`matplotlib`, and for :term:`Python` itself.

Fixes
-----

- `190 <https://github.com/usnistgov/fipy/issues/190>`_:
"matplotlib: list index out of range" when no title given, but only
sometimes
- `182 <https://github.com/usnistgov/fipy/issues/182>`_:
`~binOp` doesn't work on branches/version-2_0
- `180 <https://github.com/usnistgov/fipy/issues/180>`_:
broken arithmetic face to cell distance calculations
- `179 <https://github.com/usnistgov/fipy/issues/179>`_:
`easy_install` instructions for Mac OS X are broken
- `177 <https://github.com/usnistgov/fipy/issues/177>`_:
broken `setuptools` url with python 2.6
- `169 <https://github.com/usnistgov/fipy/issues/169>`_:
The FiPy webpage seems to be broken on Internet Explorer
- `156 <https://github.com/usnistgov/fipy/issues/156>`_:
update the mayavi viewer to use mayavi 2
- `153 <https://github.com/usnistgov/fipy/issues/153>`_:
Switch documentation to use `:math:` directive

.. [almost] Only two examples from :term:`FiPy` 2.0 fail when run with
:term:`FiPy` 2.1:

* :file:`examples/phase/symmetry.py` fails because
:class:`~fipy.meshes.mesh.Mesh` no longer provides a
``getCells`` method. The mechanism
for enforcing symmetry in the updated example is both clearer and
faster.

* :mod:`examples.levelSet.distanceFunction.circle` fails because of a
change in the comparison of masked values.

Both of these are subtle issues unlikely to affect very many
:term:`FiPy` users.

--------------------------

--------------------------

Fixes
-----

- `188 <https://github.com/usnistgov/fipy/issues/188>`_:
`SMTPSenderRefused: (553, "5.1.8 <tracmatdl-osi.org>... Domain of sender address tracmatdl-osi.org does not exist", u'"FiPy" <tracmatdl-osi.org>')`
- `184 <https://github.com/usnistgov/fipy/issues/184>`_:
`gmshExport.exportAsMesh()` doesn't work
- `183 <https://github.com/usnistgov/fipy/issues/183>`_:
FiPy 2.0.2 `LinearJORSolver.__init__` calls `Solver` rather than
`PysparseSolver`
- `181 <https://github.com/usnistgov/fipy/issues/181>`_:
Navier-Stokes again
- `151 <https://github.com/usnistgov/fipy/issues/151>`_:
update mayavi viewer to use mayavi2
- `13 <https://github.com/usnistgov/fipy/issues/13>`_:
Mesh refactor

--------------------------