sfepy windows installation

[Robert Cimrman]

Hi again!

I have finally had some success with Windows installation of SfePy. It
would be great, if someone could go through the following steps
(assuming EPD [1] is installed) and report any problems (or an unlikely
success) on the way:

1. install
http://msysgit.googlecode.com/files/Git-1.6.3.2-preview20090608.exe

2. in a terminal, try typing 'git'

3. if git works, try in some directory:

git clone git://github.com/rc/sfepy.git
cd sfepy
git pull origin windows_install

python setup.py build_ext --inplace

4. if all is ok, try:

python simple.py input/poisson.py

5. if even this works, install paraview and look at the results in
simple.vtk, or try:

ipython -wthread

and in ipython window:

%run postproc.py simple.vtk

cheers,
r.

[1] http://www.enthought.com/products/epd.php

[FOX]

I went through these steps but I obtained an error while running the
simulation.
When I run the file runTest.py I obtain 34 failure on 38tests.
There is a issue...I started with Python XY installed instead of EPD.
Should I retry with EPD installed or is it the same?

Bye Mattia

On 3 Lug, 12:20, Robert Cimrman <cimrm...@ntc.zcu.cz> wrote:
> Hi again!
>
> I have finally had some success with Windows installation of SfePy. It
> would be great, if someone could go through the following steps
> (assuming EPD [1] is installed) and report any problems (or an unlikely
> success) on the way:
>

> 1. installhttp://msysgit.googlecode.com/files/Git-1.6.3.2-preview20090608.exe

[Robert Cimrman]

Hi Mattia,

FOX wrote:
> I went through these steps but I obtained an error while running the
> simulation.
> When I run the file runTest.py I obtain 34 failure on 38tests.
> There is a issue...I started with Python XY installed instead of EPD.
> Should I retry with EPD installed or is it the same?

I have tested it with EPD only, so I cannot tell right now what might
have gone wrong - maybe some critical package that is a part of EPD is
missing. Could you, please, re-run runTest.py with the --debug flag, i.e.
runTest.py --debug
or
runTest.py --debug <failing test file>
and send the output here?

thanks,
r.

[FOX]

Hi Robert

On 20 Lug, 13:19, Robert Cimrman <cimrm...@ntc.zcu.cz> wrote:
> Hi Mattia,
>
> FOX wrote:
> > I went through these steps but I obtained an error while running the
> > simulation.
> > When I run the file runTest.py I obtain 34 failure on 38tests.
> > There is a issue...I started with Python XY installed instead of EPD.
> > Should I retry with EPD installed or is it the same?
>
> I have tested it with EPD only, so I cannot tell right now what might
> have gone wrong - maybe some critical package that is a part of EPD is
> missing.

Ok you were right, in fact the error was due to the missing of the
pyparsing module.

Could you, please, re-run runTest.py with the --debug flag, i.e.
> runTest.py --debug
> or
> runTest.py --debug <failing test file>
> and send the output here?

running
runTest.py --debug
the test stalled and gave this output

Thanks Mattia

Executing Python(x,y) 2.1.14 profile startup script: default.py
Loading NumPy
Loading SciPy
Importing all NumPy functions, modules and classes
Logging to C:\Programmi\pythonxy\logs\2009-07-20.py

Python 2.5.4 (r254:67916, Dec 23 2008, 15:10:54) [MSC v.1310 32 bit
(Intel)]
Type "copyright", "credits" or "license" for more information.

IPython 0.9.1 -- An enhanced Interactive Python.
? -> Introduction and overview of IPython's features.
%quickref -> Quick reference.
help -> Python's own help system.
object? -> Details about 'object'. ?object also works, ?? prints
more.

IPython profile: xy

Welcome to pylab, a matplotlib-based Python environment.
For more information, type 'help(pylab)'.

In [1]: cd C:\\Documents and Settings\\mattia\\sfepy
C:\Documents and Settings\mattia\sfepy

In [2]: run ./runTests.py --debug
<<< directory: tests, test files: 32
<<< tests\test_assembling.py
sfepy: left over: ['TestCommon', 'nm', '_filename', 'match_y_line',
'__builtins__', 'match_z_plane', 'match_x_line', 'match_x_plane',
'__doc__', 'match_grid_line', '__file__', 'match_y_plane',
'match_grid_plane', 'match_z_line', 'set_accuracy', '__name__', 'eps']
sfepy: reading mesh (database\tests\plane.mesh)...
sfepy: ...done in 0.01 s
sfepy: setting up domain edges...
sfepy: ...done in 0.00 s
sfepy: creating regions...
sfepy: leaf Right region_Right
sfepy: leaf LeftStrip region_LeftStrip
sfepy: leaf Omega region_Omega
sfepy: leaf RightStrip region_RightStrip
sfepy: leaf Left region_Left
sfepy: op RightFix region_RightFix
sfepy: op LeftFix region_LeftFix
sfepy: ...done in 0.05 s
sfepy: equation "eq":
sfepy: dw_diffusion.i1.Omega( m.K, q, p ) = 0
sfepy: setting up dof connectivities...
sfepy: ...done in 0.00 s
sfepy: describing geometries...
sfepy: ...done in 0.02 s
sfepy: using solvers:
nls: newton
ls: ls
sfepy: matrix shape: (460, 460)
sfepy: assembling matrix graph...
sfepy: ...done in 0.00 s
sfepy: matrix structural nonzeros: 4938 (2.33e-002% fill)
sfepy: updating materials...
sfepy: m
sfepy: ...done in 0.00 s
C:\Python25\lib\site-packages\scipy\linsolve\__init__.py:4:
DeprecationWarning:
scipy.linsolve has moved to scipy.sparse.linalg.dsolve
warn('scipy.linsolve has moved to scipy.sparse.linalg.dsolve',
DeprecationWarning)
sfepy: nls: iter: 0, residual: 8.615423e+001 (rel: 1.000000e+000)
sfepy: rezidual: 0.00 [s]
sfepy: solve: 0.07 [s]
sfepy: matrix: 0.01 [s]
sfepy: nls: iter: 1, residual: 9.988265e-014 (rel: 1.159347e-015)
>>> test instance prepared (3 test(s))
C:\Documents and Settings\mattia\sfepy\sfepy\fem\extmods\geometry.py:
174: DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
return _geometry.VolumeGeometry_variable(self, *args)
C:\Documents and Settings\mattia\sfepy\sfepy\fem\extmods\geometry.py:
174: DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use
PyArray_NewFromDescr.
return _geometry.VolumeGeometry_variable(self, *args)
... de_grad: min, max: -2126.70755439 2989.19932709
... dq_grad: min, max: -2126.70755439 2989.19932709
... ||de mode - dq mode||: 0.000000e+000
+++ test_dq_de: ok
... with unknown: d_surface_integrate.isurf.Left( p ), value: 1.0, ok:
True
... with parameter: d_surface_integrate.isurf.Left( r ), value: 1.0,
ok: True
+++ test_surface_evaluate: ok
sfepy: equation "eq":
sfepy: dw_diffusion.i1.Omega( m.K, q, p ) = 0
sfepy: setting up dof connectivities...
sfepy: ...done in 0.00 s
sfepy: describing geometries...
sfepy: ...done in 0.00 s
sfepy: matrix shape: (460, 460)
sfepy: assembling matrix graph...
sfepy: ...done in 0.00 s
sfepy: matrix structural nonzeros: 4938 (2.33e-002% fill)
sfepy: updating materials...
sfepy: m
sfepy: ...done in 0.00 s
sfepy: setting up dof connectivities...
sfepy: ...done in 0.00 s
sfepy: updating materials...
sfepy: m
sfepy: ...done in 0.00 s
sfepy: setting up dof connectivities...
sfepy: ...done in 0.00 s
sfepy: matrix shape: (493, 493)
sfepy: assembling matrix graph...
sfepy: ...done in 0.00 s
sfepy: matrix structural nonzeros: 5359 (2.20e-002% fill)
sfepy: updating materials...
sfepy: m
sfepy: ...done in 0.00 s
... ||vector mode - matrix mode||: 1.099701e-014
+++ test_vector_matrix: ok
>>> all passed in 0.18 s
<<< tests\test_base.py
sfepy: left over: ['assert_', '_filename', '__builtins__', '__name__',
'__file__', 'TestCommon', '__doc__']
>>> test instance prepared (2 test(s))
+++ test_struct_add: ok
+++ test_struct_i_add: ok
>>> all passed in 0.00 s
<<< tests\test_elasticity_small_strain.py
sfepy: left over: ['__builtins__', '__file__', 'all_your_bases',
'equations_general', '__name__', 'equations_iso', 'TestCommon',
'_filename', 'filename_meshes', 'get_pars', '__doc__']
>>> test instance prepared (2 test(s))
[('test_get_solution', <bound method Test.test_get_solution of Test>),
('test_linear_terms', <bound method Test.test_linear_terms of Test>)]
... mesh: ../database/kostka_medium_tetra.mesh, base: {'Omega':
'3_4_P1'}
... isotropic
sfepy: reading mesh (database\kostka_medium_tetra.mesh)...
sfepy: ...done in 0.05 s
sfepy: setting up domain edges...
sfepy: ...done in 0.03 s
sfepy: setting up domain faces...
sfepy: ...done in 0.03 s
sfepy: creating regions...
sfepy: leaf Top region_Top
sfepy: leaf Omega region_Omega
sfepy: leaf Bottom region_Bottom
sfepy: ...done in 0.06 s
sfepy: equation "balance_of_forces":
sfepy: dw_lin_elastic_iso.i1.Omega( solid.lame, v, u )
= dw_point_lspring.i1.Bottom( spring.pars, v, u )
sfepy: setting up dof connectivities...
sfepy: ...done in 0.00 s
sfepy: describing geometries...
sfepy: ...done in 0.03 s
sfepy: using solvers:
nls: newton
ls: ls
sfepy: matrix shape: (1287, 1287)
sfepy: assembling matrix graph...
sfepy: ...done in 0.03 s
sfepy: matrix structural nonzeros: 45993 (2.78e-002% fill)
sfepy: updating materials...
sfepy: solid
sfepy: spring
sfepy: ...done in 0.00 s
sfepy: nls: iter: 0, residual: 1.389999e+000 (rel: 1.000000e+000)
sfepy: rezidual: 0.01 [s]
sfepy: solve: 1.66 [s]
sfepy: matrix: 0.04 [s]
sfepy: nls: iter: 1, residual: 1.078579e-015 (rel: 7.759569e-016)
... converged: True
... general
sfepy: reading mesh (database\kostka_medium_tetra.mesh)...
sfepy: ...done in 0.05 s
sfepy: setting up domain edges...
sfepy: ...done in 0.03 s
sfepy: setting up domain faces...
sfepy: ...done in 0.03 s
sfepy: creating regions...
sfepy: leaf Top region_Top
sfepy: leaf Omega region_Omega
sfepy: leaf Bottom region_Bottom
sfepy: ...done in 0.06 s
sfepy: equation "balance_of_forces":
sfepy: dw_lin_elastic.i1.Omega( solid.Dijkl, v, u )
= dw_point_lspring.i1.Bottom( spring.pars, v, u )
sfepy: setting up dof connectivities...
sfepy: ...done in 0.00 s
sfepy: describing geometries...
sfepy: ...done in 0.03 s
sfepy: using solvers:
nls: newton
ls: ls
sfepy: matrix shape: (1287, 1287)
sfepy: assembling matrix graph...
sfepy: ...done in 0.03 s
sfepy: matrix structural nonzeros: 45993 (2.78e-002% fill)
sfepy: updating materials...
sfepy: solid
sfepy: spring
sfepy: ...done in 0.00 s
sfepy: nls: iter: 0, residual: 1.389999e+000 (rel: 1.000000e+000)
sfepy: rezidual: 0.02 [s]
sfepy: solve: 1.68 [s]
sfepy: matrix: 0.03 [s]
sfepy: nls: iter: 1, residual: 1.078966e-015 (rel: 7.762353e-016)
... converged: True
... mesh: ../database/kostka_medium_tetra.mesh, base: {'Omega':
'3_4_P2'}
... isotropic
sfepy: reading mesh (database\kostka_medium_tetra.mesh)...
sfepy: ...done in 0.05 s
sfepy: setting up domain edges...
sfepy: ...done in 0.03 s
sfepy: setting up domain faces...
sfepy: ...done in 0.03 s
sfepy: creating regions...
sfepy: leaf Top region_Top
sfepy: leaf Omega region_Omega
sfepy: leaf Bottom region_Bottom
sfepy: ...done in 0.06 s
sfepy: equation "balance_of_forces":
sfepy: dw_lin_elastic_iso.i1.Omega( solid.lame, v, u )
= dw_point_lspring.i1.Bottom( spring.pars, v, u )
sfepy: setting up dof connectivities...
sfepy: ...done in 0.01 s
sfepy: describing geometries...
sfepy: ...done in 0.09 s
sfepy: using solvers:
nls: newton
ls: ls
sfepy: matrix shape: (8622, 8622)
sfepy: assembling matrix graph...
sfepy: ...done in 0.29 s
sfepy: matrix structural nonzeros: 647448 (8.71e-003% fill)
sfepy: updating materials...
sfepy: solid
sfepy: spring
sfepy: ...done in 0.00 s
sfepy: nls: iter: 0, residual: 1.924264e+000 (rel: 1.000000e+000)

[Robert Cimrman]

FOX wrote:
> Hi Robert
>
> On 20 Lug, 13:19, Robert Cimrman <cimrm...@ntc.zcu.cz> wrote:
>> Hi Mattia,
>>
>> FOX wrote:
>>> I went through these steps but I obtained an error while running the
>>> simulation.
>>> When I run the file runTest.py I obtain 34 failure on 38tests.
>>> There is a issue...I started with Python XY installed instead of EPD.
>>> Should I retry with EPD installed or is it the same?
>> I have tested it with EPD only, so I cannot tell right now what might
>> have gone wrong - maybe some critical package that is a part of EPD is
>> missing.
>
> Ok you were right, in fact the error was due to the missing of the
> pyparsing module.

Good, now it seems to work (somehow) :)

> Could you, please, re-run runTest.py with the --debug flag, i.e.
>> runTest.py --debug
>> or
>> runTest.py --debug <failing test file>
>> and send the output here?
>
> running
> runTest.py --debug
> the test stalled and gave this output

It is (probably) normal situation in case you do not have umfpack
installed. One of the problems in [1] that stalls is relatively big in
terms of the matrix size, and the default superlu solver (that is in
scipy) sucks.

[1] tests\test_elasticity_small_strain.py

I have seen this happen, but had no patience to wait - maybe it will
finish one day ;)

To fix this, either install umfpack and the umfpack scikit (I can assist
you), or skip this test - the code seems to work otherwise - and try
some files in input/ manually (using simply.py) - there are also
examples how to change the solver, to, e.g., scipy iterative solvers in
ests\test_linear_solvers.

hth,
r.

[Mattia Assanelli]

To fix this, either install umfpack and the umfpack scikit (I can assist
 you),

It would be great! Please assist me...

Thanks Mattia

 

[Robert Cimrman]

Mattia Assanelli wrote:
>> To fix this, either install umfpack and the umfpack scikit (I can assist
>> you),
>
>
> It would be great! Please assist me...

Unfortunately, I do not have access to a windows box right now, but:

1. http://www.cise.ufl.edu/research/sparse/umfpack/

- try to install umfpack as described above - might be hard

2. http://scikits.appspot.com/umfpack

- install the scikit (= optional scipy module) - should be easy

Anyway, I am now hanging on #sfepy at freenode IRC channel, so ask there
if problems occur. (Not sure what to use as IRC client on windows, maybe
[1]?)

cheers,
r.

[1] http://www.mirc.com/

[Mattia Assanelli]

2009/7/20 Robert Cimrman <cimr...@ntc.zcu.cz>

Mattia Assanelli wrote:
>> To fix this, either install umfpack and the umfpack scikit (I can assist
>>  you),
>
>
> It would be great! Please assist me...

Unfortunately, I do not have access to a windows box right now, but:

1. http://www.cise.ufl.edu/research/sparse/umfpack/

- try to install umfpack as described above - might be hard

I tried installing umfpack, but I get an error while compiling UFConfig

gcc -O3 -mno-cygwin  -I../Include -I../../AMD/Include -I../../UFconfig -o umfpack_di_demo um
_demo.c ../Lib/libumfpack.a ../../AMD/Lib/libamd.a -lblas -lgfortran -lgfortranbegin -lg2c
c:\MinGW\bin\..\lib\gcc\mingw32\3.4.5\..\..\..\..\mingw32\bin\ld.exe: cannot find -lblas

Should I install lblas? Can you suggest me any tip?

[Robert Cimrman]

Mattia Assanelli wrote:
> 2009/7/20 Robert Cimrman <cimr...@ntc.zcu.cz>
>
>> Mattia Assanelli wrote:
>>>> To fix this, either install umfpack and the umfpack scikit (I can assist
>>>> you),
>>>
>>> It would be great! Please assist me...
>> Unfortunately, I do not have access to a windows box right now, but:
>>
>> 1. http://www.cise.ufl.edu/research/sparse/umfpack/
>>
>> - try to install umfpack as described above - might be hard
>
>
> I tried installing umfpack, but I get an error while compiling UFConfig
>
> gcc -O3 -mno-cygwin -I../Include -I../../AMD/Include -I../../UFconfig -o
> umfpack_di_demo um
> _demo.c ../Lib/libumfpack.a ../../AMD/Lib/libamd.a -lblas -lgfortran
> -lgfortranbegin -lg2c
> c:\MinGW\bin\..\lib\gcc\mingw32\3.4.5\..\..\..\..\mingw32\bin\ld.exe: cannot
> find -lblas
>
> Should I install lblas? Can you suggest me any tip?

Yes, you can use e.g. ATLAS [1], or GotoBLAS [2] but I have no
experience with installing it on windows. One option would be IMHO to
search numpy/scipy mailing list archives (or ask there), as those
packages rely on blas/lapack too to be fast.

[1] http://math-atlas.sourceforge.net/
[2] http://www.tacc.utexas.edu/resources/software/#blas

Sorry for the inconvenience but you see that the default direct solver
in scipy does not work well.

r.