Rewritten calibration functions

[ilario]

Trying to measure small currents with a 1 Ohm (for currents up to 80 mA) or a 10 Ohm (for currents up to 4 mA) shunt resistor I created new calibration functions, following the instructions in the code comments and in the INA219 datasheet but they were all wrong.
For example, following the instructions I ended up using a calibration factor of 41, that resulted in very poor resolution.

So, I would consider that the instructions are confusing, and I rewrote them.

As you can see, the logic I propose is very different:

1. for gain/8, use always 4096 as calibration factor (unless you are calibrating versus a reliably measured current). For other gains, 4096 is also a good choice (but also 8192 and other values would be equally as good, I just used 4096 for simplicity).
2. choose the gain that better suits your needs
3. consider replacing the shunt resistor when no gain is a good fit

Please check if the code works well also in other cases (I checked only the last two functions with shunt resistors of 1 and 10 Ohm and for currents far from the saturation ones).

[ilario]

Pylint was complaining about too many instance attributes because I used

self.calibration = 4096

instead of the previously present

self._cal_value = 4096
self._raw_calibration = self._cal_value

As I don't think I made the situation worse, I silenced the "too many instance attributes" error.
If you consider I should address this, please advise how.

[FoamyGuy]

I merged main and resolved conflicts created by newer commits.

This looks good to me. The updated equations appear to match the information in the the data sheet to me. I don't have hardware to test.

We can always revert or update as needed if people notice issues with the new version.