UCD stat quantities P/S question

[Sebastien Derriere]

   Hi Mark,

   I'd go along the lines of François on this topic.
   The rule-of-thumb for UCDs is that "the first word carries most of
the meaning". This can be tricky in some situations, and one could
argue on the semantics, but still.
   The mean value of a velocity is still a velocity, hence
phys.veloc;stat.mean  makes sense.

   A standard deviation of velocity, even if expressed in km/s, is
more a statistical thing than a velocity measurement, so
stat.stdev;phys.veloc

   It gets tricky when we talk about velocity ratio or differences for
example. Quite often, ratios are used to normalize a quantity to
some reference value, and arith.ratio and arith.diff have a S flag,
so suggested UCDs are

phys.veloc;arith.diff
phys.veloc;arith.ratio


Sébastien

Le 21/07/2016 11:10, Mark Taylor a écrit :
> Mireille,
>
> thank you for this response; but actually the point made by Francois
> makes a lot of sense too - if it's a mean/median/min/max of (e.g.)
> a magnitude, you can still treat it as a magnitude, but if it's
> a stdev/variance that's probably a bad idea.
>
> Now I have two answers I can choose the one I like :-)
> For now I will go with what's written in the document and
> avoid writing something like phot.bol.mag;stat.stdev.
> Cons: I can't mark something as a stdev of a known semantic quantity
> Pros: downstream software won't be encouraged to do something inappropriate
>
> Mark
>
> On Thu, 21 Jul 2016, Mireille Louys wrote:
>
>> Hi Mark ,
>>
>> Thanks for pointing this error.
>>
>> Yes all these terms should be used similarly and have a S status
>>
>> explaining the secondary position in the UCD strings.
>>
>> Sorry for that.
>>
>> Mireille
>>
>>
>> Le 20/07/2016 à 13:38, Mark Taylor a écrit :
>>> Dear semantics,
>>>
>>> I have a question about some of the stat.* UCDs.
>>>
>>> I see in the UCDlist document:
>>>
>>> S | stat.max
>>> S | stat.mean
>>> S | stat.median
>>> S | stat.min
>>>
>>> but
>>>
>>> P | stat.stdev
>>> P | stat.variance
>>>
>>> Why are max/min/mean/median designated as Secondary but
>>> stdev and variance as Primary?  I can imagine wanting to use,
>>> e.g. both
>>>
>>>      pos.parallax;stat.mean
>>>
>>> and
>>>
>>>      pos.parallax;stat.stdev
>>>
>>> but the second is apparently not allowed.
>>> I was expecting all of these to be marked S or maybe Q.
>>>
>>> Thanks for enlightenment,
>>>
>>> Mark
>>>
>>> --
>>> Mark Taylor   Astronomical Programmer   Physics, Bristol University, UK
>>> m.b.taylor at bris.ac.uk +44-117-9288776  http://www.star.bris.ac.uk/~mbt/
>>
> --
> Mark Taylor   Astronomical Programmer   Physics, Bristol University, UK
> m.b.taylor at bris.ac.uk +44-117-9288776  http://www.star.bris.ac.uk/~mbt/


-- 

    (((    Sebastien Derriere     sebastien.derriere at astro.unistra.fr
   (. .)   Observatoire de Strasbourg       Phone +33 (0) 368 852 444
  (( v ))  11, rue de l'universite        Telefax +33 (0) 368 852 432
---m-m--- F-67000 Strasbourg  France