/entry/instrument/attenuator/attenuator_transmission NX_FLOAT/NX_DIMENSIONLESS

[CV-GPhL]

Instead of
NX_NUMBER {units=NX_UNITLESS}
it should be
NX_FLOAT {units=NX_DIMENSIONLESS}

as in
http://hdrmx.medsbio.org/manual/build/html/classes/base_classes/NXattenuator.html#nxattenuator

Vote with thumbs up/down

[yayahjb]

Why NX_FLOAT rather than NX_NUMBER? Something dimensionless is required to have units. Something unitless is dimensionless, but also does not have a units attribute. If this is to have units, what are they?

…

On Tue, Nov 19, 2019 at 3:21 AM CV-GPhL ***@***.***> wrote: Instead of NX_NUMBER {units=NX_UNITLESS} it should be NX_FLOAT {units=NX_DIMENSIONLESS} as in http://hdrmx.medsbio.org/manual/build/html/classes/base_classes/NXattenuator.html#nxattenuator Vote with thumbs up/down — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#4?email_source=notifications&email_token=ABB6EAN7V7GWNMGY6UP7W53QUOOZZA5CNFSM4JO7SFJ2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4H2IRFRA>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABB6EAKERZ36WUKARIYTGIDQUOOZZANCNFSM4JO7SFJQ> .

[CV-GPhL]

I was looking at the definition of attenuator_transmission in NXattenuator, which has NX_FLOAT. Since it is defined as

The nominal amount of the beam that gets through (transmitted intensity)/(incident intensity)

it seems as if it should be NX_DIMENSIONLESS ("units for fields where the units cancel out"). But in principle I don't mind - as long as it is consistent with NXattenuator/attenuator_transmission, right?