Update api.py

[lbbrhzn]

Bugfix processing phase measurands. See #95

[github-actions]

🎉 HACS repository validator action summary 🎉
✅ The repository is not archived
✅ manifest.json file exist
✅ All required keys are present in manifest.json
✅ hacs.json has the 'name' key set
✅ The repository has a description
⚪ Ignored check: brands
✅ The repository has topics
✅ The repository has issues enabled
✅ This day ends with an 'y' (Friday)
✅ README.md exists
✅ Requirements validation
✅ Python wheels
✅ HACS load check

[lbbrhzn]

@drc38, why are positive voltages treated differently from negative?

           if metric in Measurand.voltage.value:
                sum = (
                    value[Phase.l1_n.value]
                    + value[Phase.l2_n.value]
                    + value[Phase.l3_n.value]
                )
                if sum > 0:
                    self._metrics[metric] = round(sum / 3, 1)
                else:
                    self._metrics[metric] = round(sum, 1)

[drc38]

@drc38, why are positive voltages treated differently from negative?

           if metric in Measurand.voltage.value:
                sum = (
                    value[Phase.l1_n.value]
                    + value[Phase.l2_n.value]
                    + value[Phase.l3_n.value]
                )
                if sum > 0:
                    self._metrics[metric] = round(sum / 3, 1)
                else:
                    self._metrics[metric] = round(sum, 1)

There should not be negative voltages, and the >0 was to protect against a div by 0 error.

[lbbrhzn]

  if sum > 0:
                    self._metrics[metric] = round(sum / 3, 1)
                else:
                    self._metrics[metric] = round(sum, 1)

You’re never dividing by sum though. And if sum equals zero, the metric value will be zero.

Ok if I just change it to:

self._metrics[metric] = round(sum / 3, 1)

[drc38]

😄 good spotting, I was working too late. Please update

[drc38]

Will you also update prior loop to reference self._extra_attr?

[lbbrhzn]

Will you also update prior loop to reference self._extra_attr?

I will keep that the same for now, but I do think there is some room improvement. We should aim to provide the one-phase equivalent voltage, current and power when three phases are reported. See here for some background info. Assuming the voltages Vx-n are balanced, we can simply add the currents to get the one-phase equivalent current. When line-to-line voltages are reported we should divide by sqrt(3) to get the line to neutral voltage. The line-to-neutral voltages are simply averaged.