Time & frequency units

[alecandido]

Despite we standardized the usage of floats for physical quantities, we actually did not clearly specify which are supposed to be the units.

I know they are established by convention in Qibocal, but this is also partially coming from 0.1, when you had to pass integers (and it was even explicitly checked), and forced people to use 1_000_000_000 instead of 1e9 to specify a GHz.

Despite floats are convenient enough that allow to use 6.34e9 and 10e-9 (for GHz and ns), in principle we may get rid of a bunch of exponents just using GHz and ns as the basic units, which are the most natural for us.

In any case, this is not mandatory, and input values still span order of magnitudes, such that you will still have to write some 1e4/1e-3 somewhere. And in any case, the update should be postponed to Qibolab 0.3, because the present QM driver is fixing a gauge, and changing would be breaking.

What would you prefer to do? @stavros11 @hay-k @andrea-pasquale @Edoardo-Pedicillo

[stavros11]

What would you prefer to do? @stavros11 @hay-k @andrea-pasquale @Edoardo-Pedicillo

Using GHz and ns sounds reasonable to me, for our current platforms.

[andrea-pasquale]

Fine by me as well