[NumPy-Tickets] [NumPy] #1913: PyUFunc_RegisterLoopForType doesn't work for custom data types

NumPy Trac numpy-tickets@scipy....
Thu Jul 21 16:06:38 CDT 2011

#1913: PyUFunc_RegisterLoopForType doesn't work for custom data types
 Reporter:  jvporter  |       Owner:  somebody   
     Type:  defect    |      Status:  new        
 Priority:  normal    |   Milestone:  Unscheduled
Component:  Other     |     Version:  1.6.0      
 Keywords:            |  
 I upgraded from NumPy 1.5 to 1.6, and found that ufuncs on my custom array
 data type stopped working (specifically == and !=). I don't think I'm
 doing anything wrong, but it's possible that something about the API has
 changed here. For reference, the code where I'm attempting this looks like
 so (minus some preprocessor magic):

     HANDLE_NPYTYPE = PyArray_RegisterDataType(descr);

     types[0] = types[1] = HANDLE_NPYTYPE; types[2] = NPY_BOOL;
     PyUFunc_RegisterLoopForType(eq, HANDLE_NPYTYPE,

 If you need more context, the full source file is here:

Ticket URL: <http://projects.scipy.org/numpy/ticket/1913>
NumPy <http://projects.scipy.org/numpy>
My example project

More information about the NumPy-Tickets mailing list