Re: [Ffc] ferari dependency
On Wed, Feb 10, 2010 at 08:50:56AM +0100, Kristian Oelgaard wrote: > > > On 10 February 2010 02:15, Garth N. Wells wrote: > >FFC now depends on ferari, > > > > File > >"/usr/local/lib/python2.6/dist-packages/ffc/tensor/tensoroptimization.py", > >line 17, in optimize_integral_ir > > from ferari import binary > >ImportError: No module named ferari > > How did you find this bug? For me it only shows if I compile with -r tensor > -O, I'm surprised that you're using that combination :) > Anyway, I pushed a fix for this. > > Kristian Good that you fixed it. The import was actually inside try clause with a suitable warning if FErari is not available, but I had some problems which led me to adding an import before the try for debugging. I forgot to remove it. FErari should be an optional dependency. -- Anders signature.asc Description: Digital signature ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp
Re: [Ffc] ferari dependency
Kristian Oelgaard wrote: > > > On 10 February 2010 02:15, Garth N. Wells wrote: >> FFC now depends on ferari, >> >> File >> "/usr/local/lib/python2.6/dist-packages/ffc/tensor/tensoroptimization.py", >> >> line 17, in optimize_integral_ir >>from ferari import binary >> ImportError: No module named ferari > > How did you find this bug? For me it only shows if I compile with -r > tensor -O, I'm surprised that you're using that combination :) FFC is very clever these days. It chooses quadrature for the 'real' solve, and then the tensor representation for some projections. Garth > Anyway, I pushed a fix for this. > > Kristian > >> I don't think that it should. >> >> Garth >> >> ___ >> Mailing list: https://launchpad.net/~ffc >> Post to : ffc@lists.launchpad.net >> Unsubscribe : https://launchpad.net/~ffc >> More help : https://help.launchpad.net/ListHelp >> > ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp
Re: [Ffc] ferari dependency
On Wed, Feb 10, 2010 at 08:55:29AM +, Garth N. Wells wrote: > > > Kristian Oelgaard wrote: > > > > > > On 10 February 2010 02:15, Garth N. Wells wrote: > >> FFC now depends on ferari, > >> > >> File > >> "/usr/local/lib/python2.6/dist-packages/ffc/tensor/tensoroptimization.py", > >> > >> line 17, in optimize_integral_ir > >>from ferari import binary > >> ImportError: No module named ferari > > > > How did you find this bug? For me it only shows if I compile with -r > > tensor -O, I'm surprised that you're using that combination :) > > FFC is very clever these days. It chooses quadrature for the 'real' > solve, and then the tensor representation for some projections. :-) -- Anders signature.asc Description: Digital signature ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp
[Ffc] form object not iterable
hi, i just pulled from the ffc repository and i get this error when compiling my forms with '--debug': Compiler stage 4.1: Generating additional wrapper code -- Traceback (most recent call last): File "/scratch-second/priesen/FEniCS/build/bin/ffc", line 167, in sys.exit(main(sys.argv[1:])) File "/scratch-second/priesen/FEniCS/build/bin/ffc", line 149, in main compile_form(ufd.forms, ufd.object_names, prefix, parameters) File "/scratch-second/priesen/FEniCS/build/lib/python2.5/site-packages/ffc/compiler.py", line 155, in compile_form wrapper_code = generate_wrapper_code(analysis, prefix, parameters) File "/scratch-second/priesen/FEniCS/build/lib/python2.5/site-packages/ffc/wrappers.py", line 47, in generate_wrapper_code for (i, (form, form_data)) in enumerate(form_and_data): TypeError: 'Form' object is not iterable regards, patrick ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp
Re: [Ffc] form object not iterable
It seems that you have forgotten to update UFL. Mehdi On Wed, 2010-02-10 at 13:19 +0100, Patrick Riesen wrote: > hi, i just pulled from the ffc repository and i get this error when > compiling my forms with '--debug': > >Compiler stage 4.1: Generating additional wrapper code >-- > Traceback (most recent call last): >File "/scratch-second/priesen/FEniCS/build/bin/ffc", line 167, in > > sys.exit(main(sys.argv[1:])) >File "/scratch-second/priesen/FEniCS/build/bin/ffc", line 149, in main > compile_form(ufd.forms, ufd.object_names, prefix, parameters) >File > "/scratch-second/priesen/FEniCS/build/lib/python2.5/site-packages/ffc/compiler.py", > > line 155, in compile_form > wrapper_code = generate_wrapper_code(analysis, prefix, parameters) >File > "/scratch-second/priesen/FEniCS/build/lib/python2.5/site-packages/ffc/wrappers.py", > > line 47, in generate_wrapper_code > for (i, (form, form_data)) in enumerate(form_and_data): > TypeError: 'Form' object is not iterable > > > regards, > patrick > > ___ > Mailing list: https://launchpad.net/~ffc > Post to : ffc@lists.launchpad.net > Unsubscribe : https://launchpad.net/~ffc > More help : https://help.launchpad.net/ListHelp ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp
[Ffc] Bug in generating wrapper code
When compiling with -l dolfin, the DOLFIN Poisson demo does not get the named f, g coefficients but instead w0 and w1. Don't know why yet and I won't be able to debug until later. Have teaching now. -- Anders signature.asc Description: Digital signature ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp
Re: [Ffc] Bug in generating wrapper code
On 10 February 2010 13:39, Anders Logg wrote: When compiling with -l dolfin, the DOLFIN Poisson demo does not get the named f, g coefficients but instead w0 and w1. Don't know why yet and I won't be able to debug until later. Have teaching now. I think I fixed this in analysis.py, just added the dictionary object_names as argument to preprocess() :) Kristian -- Anders -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) iEYEARECAAYFAktyqPEACgkQTuwUCDsYZdGAswCeNdyQasC36o7U2OVqN03wNx2z aJcAn2w9v834h5jDs7e/Peh89jWnveFs =zHdx -END PGP SIGNATURE- ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp signature.asc Description: OpenPGP digital signature ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp
Re: [Ffc] form object not iterable
hello mehdi, i did update all packages, including UFL . if i do not use the option '-l dolfin' it works. patrick Mehdi Nikbakht wrote: It seems that you have forgotten to update UFL. Mehdi On Wed, 2010-02-10 at 13:19 +0100, Patrick Riesen wrote: hi, i just pulled from the ffc repository and i get this error when compiling my forms with '--debug': Compiler stage 4.1: Generating additional wrapper code -- Traceback (most recent call last): File "/scratch-second/priesen/FEniCS/build/bin/ffc", line 167, in sys.exit(main(sys.argv[1:])) File "/scratch-second/priesen/FEniCS/build/bin/ffc", line 149, in main compile_form(ufd.forms, ufd.object_names, prefix, parameters) File "/scratch-second/priesen/FEniCS/build/lib/python2.5/site-packages/ffc/compiler.py", line 155, in compile_form wrapper_code = generate_wrapper_code(analysis, prefix, parameters) File "/scratch-second/priesen/FEniCS/build/lib/python2.5/site-packages/ffc/wrappers.py", line 47, in generate_wrapper_code for (i, (form, form_data)) in enumerate(form_and_data): TypeError: 'Form' object is not iterable regards, patrick ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp
Re: [Ffc] form object not iterable
i just pulled again from the FFC repo, i works now :-) thank you, patrick Patrick Riesen wrote: hello mehdi, i did update all packages, including UFL . if i do not use the option '-l dolfin' it works. patrick Mehdi Nikbakht wrote: It seems that you have forgotten to update UFL. Mehdi On Wed, 2010-02-10 at 13:19 +0100, Patrick Riesen wrote: hi, i just pulled from the ffc repository and i get this error when compiling my forms with '--debug': Compiler stage 4.1: Generating additional wrapper code -- Traceback (most recent call last): File "/scratch-second/priesen/FEniCS/build/bin/ffc", line 167, in sys.exit(main(sys.argv[1:])) File "/scratch-second/priesen/FEniCS/build/bin/ffc", line 149, in main compile_form(ufd.forms, ufd.object_names, prefix, parameters) File "/scratch-second/priesen/FEniCS/build/lib/python2.5/site-packages/ffc/compiler.py", line 155, in compile_form wrapper_code = generate_wrapper_code(analysis, prefix, parameters) File "/scratch-second/priesen/FEniCS/build/lib/python2.5/site-packages/ffc/wrappers.py", line 47, in generate_wrapper_code for (i, (form, form_data)) in enumerate(form_and_data): TypeError: 'Form' object is not iterable regards, patrick ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp
Re: [Ffc] Bug in generating wrapper code
On Wed, Feb 10, 2010 at 02:25:42PM +0100, Kristian Oelgaard wrote: > > > On 10 February 2010 13:39, Anders Logg wrote: > >When compiling with -l dolfin, the DOLFIN Poisson demo does not get > >the named f, g coefficients but instead w0 and w1. Don't know why yet > >and I won't be able to debug until later. Have teaching now. > > I think I fixed this in analysis.py, just added the dictionary object_names > as argument to preprocess() :) > > Kristian Great! -- Anders signature.asc Description: Digital signature ___ Mailing list: https://launchpad.net/~ffc Post to : ffc@lists.launchpad.net Unsubscribe : https://launchpad.net/~ffc More help : https://help.launchpad.net/ListHelp
