[Numpy-discussion] default float type of array not accepted by SWIG wrapped C functions
Thu Jan 22 10:45:51 CST 2009
I have a SWIG wrapped C library that uses 32bit floating point arrays,
using the numpy.i typemapping system for passing the arrays. For
every array that I make, I have to convert it using astype('float32'),
else python complains that I tried to pass a double-precision array.
Is there any way to set the default floating point precision to 32bit,
in python or in the SWIG interface file?
More information about the Numpy-discussion