[SCHEMA] (for v2.0) Should calibration["measurement_type_id"] be required?

[dancasey-ie]

Should calibration["measurement_type_id"] be required?

Both because sensors can have multiple measurement_types that can be calibrated separately that need to be differentiated and the current common measurement_type definition does not accept null values.

Longer explanation:
calibration["measurement_type_id"] shares the same definition as measurement_point["measurement_type_id"]. The definition does not allow null values. measurement_point["measurement_type_id"] is a required field so there is no issue, but calibration["measurement_type_id"] is currently an optional field so can cause errors if left blank.

[abohara]

Hi @dancasey-ie

I agree conceptually that a sensor can have multiple measurement and hence it is possible to have multiple calibrations for the same sensor pertaining to this different measurements. Do you have an actual use case we can use as an example to discuss further ?

cc: @stephenholleran @heikowestermann

[heikowestermann]

Hey @dancasey-ie @abohara @stephenholleran,
I think measurement_type_id should be required as otherwise the measurand/measurement unit cannot be correctly identified. An example for multiple calibrations per instrument would be an ultrasonic anemometer which outputs the wind speed and the wind direction and both can be calibrated according to IEC61400-12-1 (IEC61400-50-1). Since the calibration key can also be an array, I think this is currently also supported by the schema.

[abohara]

Thanks for the clarification @heikowestermann and @dancasey-ie

I misunderstood the request. Agree that it is a required item and should be marked as such in the schema.

[stephenholleran]

Thanks @heikowestermann @abohara and @dancasey-ie.

Making it required will be a breaking change. Not ideal but I think the next release is going to have a few anyway.

I'll make the change.

[dancasey-ie]

Thanks all @heikowestermann @abohara and @stephenholleran .