Possibly related to #86. When passing a numpy float64 to a dll library using PythonNet, the object type changes (not sure to what type).
I have tried to reproduce the error on the dll directly, by passing the same value with different types and it works as expected. Any ideas?