You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I upgraded my numpy to the latest version in pip (1.9) and started getting errors when importing dolfin like these guys ("ImportError: No module named oldnumeric"): spectralpython/spectral#13
I downgraded to numpy 1.8.2 and my dolfin went back to working.
Apparently oldnumeric has been removed, and my Fenics depended on it.
I checked the install scripts, and it looks like the run.ubuntu.sh ones use the package manager numpy, which is out of date.
I think on the Mac, the python in the devtools comes with an outdated numpy as well that seems to work.
Anyway, I think we're lucking out of hitting this problem now but maybe it's important later. I think Ubuntu 14.04 still uses numpy 1.8.1 so we're still safe.
The text was updated successfully, but these errors were encountered:
For an OSX 10.9.5 install with FEnICs 1.4.0 (version specifically requested) this fixed the following error:
"Exception: Unable to import the Python Scientific module required by FIAT.
Consider installing the package python-scientific."
It is still a bug for OSX 10.9 with FEnICs 1.4.0 - your post gave me the clue on how to get dolfin to work with the FEnICs 1.4.0 OSX binary. Thanks again!!!
I upgraded my numpy to the latest version in pip (1.9) and started getting errors when importing dolfin like these guys ("ImportError: No module named oldnumeric"):
spectralpython/spectral#13
I downgraded to numpy 1.8.2 and my dolfin went back to working.
Apparently oldnumeric has been removed, and my Fenics depended on it.
I checked the install scripts, and it looks like the run.ubuntu.sh ones use the package manager numpy, which is out of date.
I think on the Mac, the python in the devtools comes with an outdated numpy as well that seems to work.
Anyway, I think we're lucking out of hitting this problem now but maybe it's important later. I think Ubuntu 14.04 still uses numpy 1.8.1 so we're still safe.
The text was updated successfully, but these errors were encountered: