Good:
>>> inf = float('inf')
>>> mpmath.bernpoly(3, inf)
mpf('+inf')
>>> mpmath.bernpoly(4, inf)
mpf('+inf')
Now how about complex with 0 imag part? This seems ok:
>>> bernpoly(4, complex(inf, 0))
mpc(real='+inf', imag='0.0')
```
But:
```
>>> bernpoly(3, complex(inf, 0))
mpc(real='nan', imag='nan')
This isn't wrong per se if you think of complex(..., 0) as a very small negative part.
Do we care? Correct behaviour at real infinity is very nice property but maybe exactly what happens for complex(..., 0) is less important?
Good:
Now how about complex with 0 imag part? This seems ok:
This isn't wrong per se if you think of
complex(..., 0)as a very small negative part.Do we care? Correct behaviour at real infinity is very nice property but maybe exactly what happens for
complex(..., 0)is less important?