search

pele-python / mcpele

Monte Carlo and parallel tempering routines built on the pele foundation
Other
20 stars 5 forks source link

Numpy includes #52

Closed kjs73 closed 10 years ago

kjs73 commented 10 years ago

On deathstar, the setup fails if numpy includes are added as 

## Numpy header files 
numpy_lib = os.path.split(np.__file__)[0] 
numpy_include = os.path.join(numpy_lib, 'core/include')

For some reason, there it works to use instead

numpy_include = [np.get_include()]

which is taken from here: http://stackoverflow.com/questions/14657375/cython-fatal-error-numpy-arrayobject-h-no-such-file-or-directory

On my desktop, e.g., both work and give the same result. For pele, wich has the same setup_with_cmake, this issue does not exist. This is very confusing to me.

js850 commented 10 years ago

I think I changed it somewhat in pele On 18 Oct 2014 15:18, "Julian Schrenk" notifications@github.com wrote:

On deathstar, the setup fails if numpy includes are added as

Numpy header files

numpy_lib = os.path.split(np.file)[0] numpy_include = os.path.join(numpy_lib, 'core/include')

For some reason, there it works to use instead

numpy_include = [np.get_include()]

which is taken from here:

http://stackoverflow.com/questions/14657375/cython-fatal-error-numpy-arrayobject-h-no-such-file-or-directory

On my desktop, e.g., both work and give the same result. For pele, wich has the same setup_with_cmake, this issue does not exist. This is very confusing to me.

Reply to this email directly or view it on GitHub https://github.com/pele-python/mcpele/issues/52.

js850 commented 10 years ago

You might try looking for my pele changes and making them in mcpele also. On 18 Oct 2014 18:13, "Jacob Stevenson" jstevenson131@gmail.com wrote:

I think I changed it somewhat in pele On 18 Oct 2014 15:18, "Julian Schrenk" notifications@github.com wrote:

On deathstar, the setup fails if numpy includes are added as

Numpy header files

numpy_lib = os.path.split(np.file)[0] numpy_include = os.path.join(numpy_lib, 'core/include')

For some reason, there it works to use instead

numpy_include = [np.get_include()]

which is taken from here:

http://stackoverflow.com/questions/14657375/cython-fatal-error-numpy-arrayobject-h-no-such-file-or-directory

On my desktop, e.g., both work and give the same result. For pele, wich has the same setup_with_cmake, this issue does not exist. This is very confusing to me.

Reply to this email directly or view it on GitHub https://github.com/pele-python/mcpele/issues/52.

kjs73 commented 10 years ago

OK, thanks, that worked fine!