Search code examples
pythonnumpycomcythonpywin32

python comtypes how to pass numpy float32 (singe-precision float) as scalar?

According to comtypes documentation is is possible to pass numpy arrays to comtypes objects, but how to I pass scalar values, e.g. numpy.float32(1.0)?

https://pythonhosted.org/comtypes/#numpy-interop

I get following error:

---------------------------------------------------------------------------
ArgumentError                             Traceback (most recent call last)
<ipython-input-67-9e7e5859c6a5> in <module>()
----> 1 com_obj.Update("string_name",np.float32(6.6e-6))
ArgumentError: argument 2: <type 'exceptions.TypeError'>: Cannot put 6.5999998e-06 in VARIANT
Solution

  • Passing ctypes.c_float rather than np.float32 will work. e.g. com_obj.Update("string_name", ctypes.c_float(6.6e-6)).