Styleguide

We format the code using Black.

Floating point types

They should be equivalent on the platforms we run our code on. float and np.float_ and np.float64 are all double precision floating point values which are IEEE 788 64 bits.

We write np.float64 explicitly when dealing with I/O.

We have a Float alias available to work interchangeably with float and np.float64. Other floating-point types should be avoided, except float32 for compact storage, and of course complex numbers when relevant.