<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<p>Hi all, </p>
<p>Two words on this.</p>
<p>A ) Measurements<br>
</p>
<p> I remember the discussion among authors when we added (to the
initial 1.0 list) the "measurements" term among authors of ObsCore
1.1 . Indeed I think the idea to add something like "catalog" was
given by Daniel DUrand from CADC because they had this use case
that Pat is exposing.</p>
<p>There was however a reason why we didn't choose to add the
dataproduct-type "catalog".</p>
<p> ObsCore/ObsTAP is for discovery of datasets which have time,
spatial, spectral and polarisation axes. Selection on the ObsCore
parameters is not sufficient for catalogs with plenty of other
parameters which are directly queried via TAP (or even in the
registry). So we had to find another word for these specific
tables extracted from the datasets in order to not let beleive
that ObsCore is a discovery model for any kind of catalog. Hence
"measurements"</p>
<p> HiPS was different because they are really creating HiPS for
ANY KIND of catalog !!! <br>
</p>
<p> So I think we should keep "measurements" but not with the
negative definition "Generic tabular data not fitting any of the
other terms. Because<br>
of its lack of specificity, this term should generally be
avoided, and new, more precise terms should be introduced
instead" any of the others will fit I think and yes we have to
keep ascendant compatibility with obsCore. This is pretty
important for interoperability<br>
</p>
<p>B ) TimeSeries and :SED. hierarchies<br>
</p>
<p> This is maybe more critical. the definition in ObsCore which
has been reproduced in Markus' list seems to exclude TimeSeries of
spectra or Images or whatever thing which is not a varying scalar.
I don't remember the reason but where do we put spectrochronogram
then ? Do we create subtypes of spectrum or cube ?</p>
<p> The question of subtypes and "parent" terms seems to be
open. In Obscore there was no hierarchy in the terms. But in
Markus list there is sed as a child from spectrum. I think some
people complained about using dataproduct_subtype for such things
and prefer to let this subtype Obscore field available for free
strings.</p>
<p> Imagine in the future (next version of ObscORe) we decide
to have ObsCore vocabulary for dataproduct_type externally defined
as the VOcabulary list we are discussing here. This could be a way
to easilly extend ObsCore vocabulary for this specific
dataproduct_type attribute.<br>
</p>
<p> We may imagine have "spectrochronogram" and "sed" as
children elements of spectrum. "timecube" or "spectralcube" as
children from cubes. <br>
</p>
<p> This will be clear in the vocabulary page but ObsCore
will manage that at the same level in dataproduct_type (exactly
like we already have sed and spectrum in parallel today)</p>
<p> Thoughts ?</p>
<p>C ) Miscelaneous.</p>
<p> If this vocabulary is to be used in various contexts (and
indeed it is) why do we link it to SimpleDALRegExt ? Vocabularies
2.0 is proposing to manage vocabularies as endorsed notes. Why
don't we do it this way and refer it from SimpleDAL, ObsCore,
DataLink, etc ...</p>
<p> The MS discussion indeed shows we need to add new
refinements in formats conveyed by the "media types". Does one
exist for Measurement sets already ?</p>
<p>Cheers</p>
<p>François<br>
</p>
<p> <br>
</p>
<p><br>
</p>
<div class="moz-cite-prefix">Le 12/03/2020 à 16:20, Patrick Dowler a
écrit :<br>
</div>
<blockquote type="cite"
cite="mid:CAFK8nrr7JRsj_G3UH9MFWBoT_xnbCAxJra7ipcWor6Y4Bsn8yw@mail.gmail.com">
<div dir="ltr">
<div class="gmail_default">note: cross-posting reduced to DAL
and semantics</div>
<div class="gmail_default"><br>
</div>
<div class="gmail_default">In CAOM, I made an
ObsCore.dataproduyct_type base vocabulary and then added my
own custom child of measurements named <a
href="http://www.opencadc.org/caom/DataProductType#catalog"
moz-do-not-send="true">http://www.opencadc.org/caom/DataProductType#catalog</a>
(this is literally ObsCore + extensions so if this idea goes
ahead we'll have to figure out this vocab refers back to the
parent vocab). If one queries caom2 in our tap service you can
see that (fully qualified) value mixed in with ObsCore values.
In the ObsCore view I am currently limiting the rows to
exclude custom (fully qualified) terms for compliance with the
current model (1.1). In the event that:<br>
</div>
<div class="gmail_default"><br>
</div>
<div class="gmail_default">* ObsCore (1.2) dataproduct_type
values are defined by a vocabulary, and</div>
<div class="gmail_default">* providers can extend the vocabulary
with custom terms (then fully qualified rather that just the
short form when using base terms)</div>
<div class="gmail_default"><br>
</div>
<div class="gmail_default">I could remove he filtering and allow
more entries to appear in ObsCore view. If that doesn't become
possible, we would have to decide to (i) leave content as-is
or (ii) change those to measurements.</div>
<div class="gmail_default"><br>
</div>
<div class="gmail_default">Currently CADC has:</div>
<div class="gmail_default"><br>
</div>
<div class="gmail_default">select dataProductType,count(*) from
caom2.Plane group by dataProductType;<br>
dataproducttype |
count <br>
-------------------------------------------------------+----------<br>
|
15538972<br>
timeseries |
866405<br>
spectrum |
7911476<br>
cube |
202980<br>
<a
href="http://www.opencadc.org/caom2/DataProductType#catalog"
moz-do-not-send="true">http://www.opencadc.org/caom2/DataProductType#catalog</a>
| 160850<br>
eventlist |
88421<br>
image |
16076616</div>
<div class="gmail_default"><br>
</div>
<div class="gmail_default">So, we are using "measurements" but
it is not surprising that Markus didn't see it. We only have
one child term in use right now.<br>
</div>
<div class="gmail_default"><br>
</div>
<div>
<div dir="ltr" class="gmail_signature"
data-smartmail="gmail_signature">
<div dir="ltr">
<div>
<div dir="ltr">
<div>
<div>--<br>
</div>
<div>Patrick Dowler<br>
</div>
Canadian Astronomy Data Centre<br>
</div>
Victoria, BC, Canada<br>
</div>
</div>
</div>
</div>
</div>
<br>
</div>
<br>
<div class="gmail_quote">
<div dir="ltr" class="gmail_attr">On Thu, 12 Mar 2020 at 00:01,
Slava Kitaeff <<a href="mailto:slava.kitaeff@uwa.edu.au"
moz-do-not-send="true">slava.kitaeff@uwa.edu.au</a>>
wrote:<br>
</div>
<blockquote class="gmail_quote">
<div lang="EN-AU">
<div class="gmail-m_-3349331814557714389WordSection1">
<p class="MsoNormal">Hi James et all,</p>
<p class="MsoNormal"> </p>
<p class="MsoNormal">Just a quick note on that. It’s a bit
of a guess but I’m suspecting that this might be
referring to a Measurement Set (MS), which is a specific
data format (and a model) used by CASA (and now
ASKAPsoft) to store interferometric visibilities. So,
the data type is then must be “visibilities”. To confuse
things further, CASA can store continuum images and
spectral-line image-cubes in the same MS format as CASA
tables. In my view, it’d be ideal to have two things
describing a) the nature of data and b) its specific
format (including version).</p>
<p class="MsoNormal"> </p>
<p class="MsoNormal">Cheers,</p>
<p class="MsoNormal">Slava</p>
<p class="MsoNormal"><span>________________________________</span><span></span></p>
<p class="MsoNormal"> </p>
<p class="MsoNormal"><b><span>Dr Slava Kitaeff</span></b><span></span></p>
<p class="MsoNormal"><span>Australian SKA Regional Centre
Program Manager</span><span></span></p>
<p class="MsoNormal"><span> </span></p>
<p class="MsoNormal"><b><span>International Centre for
Radio Astronomy Research</span></b><span></span></p>
<p class="MsoNormal"><span>The University of Western
Australia</span><span></span></p>
<p class="MsoNormal"><span> </span></p>
<p class="MsoNormal"><b><span>Astronomy and Space Science</span></b><span></span></p>
<p class="MsoNormal"><span>CSIRO</span></p>
<p class="MsoNormal"> </p>
<p class="MsoNormal"><span>Email: </span><span><a
href="mailto:slava.kitaeff@icrar.org"
target="_blank" moz-do-not-send="true"><span>slava.kitaeff@icrar.org</span></a> </span><span>or <a
href="mailto:slava.kitaeff@csiro.au" target="_blank"
moz-do-not-send="true"><span>slava.kitaeff@csiro.au</span></a> </span><span></span></p>
<p class="MsoNormal"><span>Tel.: (+61) (0) 8 6488 7744
(ICRAR) or (+61) (0) 8 6436 8865 (CSIRO)</span><span></span></p>
<p class="MsoNormal"><span>Mob.: +61 404 297 414</span><span></span></p>
<p class="MsoNormal"><i><span> </span></i><span></span></p>
<p class="MsoNormal"><b><span>Mailing addresses:</span></b><span> </span><span></span></p>
<p class="MsoNormal"><span>M468, University of Western
Australia, 35 Stirling Highway, Crawley WA 6009,
Australia, <b>or</b></span><span></span></p>
<p class="MsoNormal"><span>CSIRO Astronomy and Space
Science, PO Box 1130, Bentley WA 6102, Australia</span><span></span></p>
<p class="MsoNormal"><span> </span></p>
<p class="MsoNormal"><span>ICRAR: Discovering the hidden
Universe through radio astronomy<b> </b></span><span></span></p>
<p class="MsoNormal"><span><a href="http://www.icrar.org"
target="_blank" moz-do-not-send="true"><span>www.icrar.org</span></a><span> </span></span><a
href="http://www.icrar.org/#subscribe" target="_blank"
moz-do-not-send="true"><span>Subscribe to our
eNewsletter</span></a><span> </span><a
href="http://twitter.com/icrar" target="_blank"
moz-do-not-send="true"><span>ICRAR on Twitter</span></a><span> </span><a
href="http://www.facebook.com/pages/ICRAR/199692286227" target="_blank"
moz-do-not-send="true"><span>ICRAR on Facebook</span></a></p>
<p class="MsoNormal"> </p>
<p class="MsoNormal"> </p>
<div>
<p class="MsoNormal"><b><span>From:
</span></b><span><<a
href="mailto:dal-bounces@ivoa.net" target="_blank"
moz-do-not-send="true">dal-bounces@ivoa.net</a>>
on behalf of "Dempsey, James (IM&T, Black
Mountain)" <a class="moz-txt-link-rfc2396E" href="mailto:James.Dempsey@csiro.au"><James.Dempsey@csiro.au></a><br>
<b>Date: </b>Wednesday, 11 March 2020 at 5:43 pm<br>
<b>To: </b>Markus Demleitner <<a
href="mailto:msdemlei@ari.uni-heidelberg.de"
target="_blank" moz-do-not-send="true">msdemlei@ari.uni-heidelberg.de</a>>,
"<a href="mailto:dal@ivoa.net" target="_blank"
moz-do-not-send="true">dal@ivoa.net</a>" <<a
href="mailto:dal@ivoa.net" target="_blank"
moz-do-not-send="true">dal@ivoa.net</a>>, "<a
href="mailto:registry@ivoa.net" target="_blank"
moz-do-not-send="true">registry@ivoa.net</a>" <<a
href="mailto:registry@ivoa.net" target="_blank"
moz-do-not-send="true">registry@ivoa.net</a>>,
"<a href="mailto:semantics@ivoa.net" target="_blank"
moz-do-not-send="true">semantics@ivoa.net</a>"
<<a href="mailto:semantics@ivoa.net"
target="_blank" moz-do-not-send="true">semantics@ivoa.net</a>><br>
<b>Subject: </b>Re: Vocabularising dataproduct_type</span></p>
</div>
<div>
<p class="MsoNormal"> </p>
</div>
<div>
<p class="MsoNormal"><span>Hi Markus,</span></p>
</div>
<div>
<p class="MsoNormal"><span> </span></p>
</div>
<div>
<p class="MsoNormal"><span>We are dealing with a lot of
catalogues as well as images, cubes etc. Currently
these have a blank dataproduct_type in the CASDA
obscore instance as there wasn't anything we could
use in v1.0. We will probably start using
measurements soon as it is better than nothing, but
it isn't a term that our community uses. Their
expectation is to see catalogue in the type, so I'd
be keen to see that as term available for use.</span></p>
</div>
<div>
<p class="MsoNormal"><span> </span></p>
</div>
<div>
<p class="MsoNormal"><span>Our catalogues are generally
source lists produced from the associated
image/cube.</span></p>
</div>
<div>
<div>
<p class="MsoNormal"><span> </span></p>
</div>
<div>
<p class="MsoNormal"><span>Cheers,</span></p>
</div>
<div id="gmail-m_-3349331814557714389Signature">
<div name="divtagdefaultwrapper">
<div>
<div>
<div>
<p class="MsoNormal"><span>James Dempsey</span></p>
</div>
<div>
<p class="MsoNormal"><span>Senior Developer</span><span></span></p>
</div>
<div>
<p class="MsoNormal"><span>Information
Services Applications</span><span></span></p>
</div>
<div>
<p class="MsoNormal"><span>CSIRO Information
Management & Technology (IM&T)</span></p>
</div>
</div>
</div>
</div>
</div>
</div>
<div class="MsoNormal">
<hr width="100%">
</div>
<div id="gmail-m_-3349331814557714389divRplyFwdMsg">
<p class="MsoNormal"><b><span>From:</span></b><span> <a
href="mailto:dal-bounces@ivoa.net" target="_blank"
moz-do-not-send="true">dal-bounces@ivoa.net</a>
<<a href="mailto:dal-bounces@ivoa.net"
target="_blank" moz-do-not-send="true">dal-bounces@ivoa.net</a>>
on behalf of Markus Demleitner <<a
href="mailto:msdemlei@ari.uni-heidelberg.de"
target="_blank" moz-do-not-send="true">msdemlei@ari.uni-heidelberg.de</a>><br>
<b>Sent:</b> Wednesday, 11 March 2020 7:48 PM<br>
<b>To:</b> <a href="mailto:dal@ivoa.net"
target="_blank" moz-do-not-send="true">dal@ivoa.net</a>
<<a href="mailto:dal@ivoa.net" target="_blank"
moz-do-not-send="true">dal@ivoa.net</a>>; <a
href="mailto:registry@ivoa.net" target="_blank"
moz-do-not-send="true">registry@ivoa.net</a> <<a
href="mailto:registry@ivoa.net" target="_blank"
moz-do-not-send="true">registry@ivoa.net</a>>;
<a href="mailto:semantics@ivoa.net" target="_blank"
moz-do-not-send="true">semantics@ivoa.net</a> <<a
href="mailto:semantics@ivoa.net" target="_blank"
moz-do-not-send="true">semantics@ivoa.net</a>><br>
<b>Subject:</b> Re: Vocabularising dataproduct_type</span>
</p>
<div>
<p class="MsoNormal"> </p>
</div>
</div>
<div>
<div>
<p class="MsoNormal">Hi Sarah,<br>
<br>
On Tue, Mar 10, 2020 at 03:58:09PM +0000, Sarah
Weissman wrote:<br>
> * Would too many things be mapped into
"Measurement" to be useful<br>
> from a discovery perspective?<br>
<br>
Given my findings yesterday (no active obscore
service right now has<br>
measurements rows), I think we shouldn't sweat this
measurements<br>
thing too much at this point -- in Registry, this
term would only be<br>
used for (SSAP, for now) services *returning* lists
of such<br>
measurements (as they would otherwise return
spectra), which seems a<br>
long shot.<br>
<br>
Normal discovery of catalogues and similar will of
course proceed as<br>
usual.<br>
<br>
However, I largely agree with Laurent:<br>
<br>
On Tue, 10 Mar 2020 17:42:43 +0100, Laurent Michel
wrote:<br>
> I would really prefer the list of
dataproduct_type to ensur an ascending<br>
> compatibility with Obscore. I believe it is
worth to keep as fas as possible a<br>
> consistance betwenn standards. This might help
in many cases to map things<br>
<br>
So, even though I think we've found that the actual
definition of<br>
measurement is, well, hard, I'm now rather convinced
we shouldn't<br>
drop it. On the other hand, when we follow Marco --<br>
<br>
On Tue, 10 Mar 2020 13:53:09 +0100, Molinaro, Marco
wrote:<br>
> Il giorno mar 10 mar 2020 alle ore 13:14 Markus
Demleitner <<br>
>> A set of derived measurements obtained
from a particular original<br>
>> dataset. The prototypical example would
be a list of sources<br>
>> extracted from an image.<br>
><br>
> Does it have to be "original dataset"
_singular_?<br>
<br>
and Laurent's previous proposal, we'd end up with:<br>
<br>
A set of derived measurements obtained from
original datasets or a<br>
catalog of sources.<br>
<br>
I'd say the "obtained from original datasets" in
there doesn't really<br>
mean much (what else could the measurements be
derived from?), so<br>
reduced it would work out to<br>
<br>
A set of derived measurements or a catalog of
sources.<br>
<br>
Which, as Sarah rightly says, is really vague and
all-encompassing.<br>
As I said, I'd normally throw out such a term on
grounds of it<br>
colliding with too many other terms without really
fitting well into<br>
a hierarchy.<br>
<br>
But again, since it's in obscore, we shouldn't drop
it. But since<br>
it, to me, seems ill-defined, perhaps we should be
frank and just say:<br>
<br>
Generic tabular data not fitting any of the other
terms. Because<br>
of its lack of specificity, this term should
generally be avoided,<br>
and new, more precise terms should be introduced
instead.<br>
<br>
Can people live with that? If we find good use
cases for<br>
"measurements" later, we still can get more precise
in this<br>
definition as long as we now say "don't use it,
really".<br>
<br>
Back to Sarah:<br>
<br>
On Tue, Mar 10, 2020 at 03:58:09PM +0000, Sarah
Weissman wrote:<br>
> * Do we need a separate category for "Model"?<br>
<br>
The future Vocabularies 2 standard is designed to
make it easy to add<br>
new terms exactly when they're actually needed. So,
when you you<br>
have an SSA or Obscore service returning such Models
instances -- or<br>
some future use of this vocabulary needs it --, we
can include it.<br>
For now, let's see if we can just stick with what
Obscore already<br>
has.<br>
<br>
> * Do you expect that more than one label would
be applied to a data<br>
> product? For example (naively) could a
"spectral image cube" be<br>
> labeled with "spectrum" and "image" and "cube"?<br>
<br>
This depends on where the terms are being used. In
obscore, only one<br>
label per row is possible, and I don't think that
can be changed in a<br>
backwards-compatible way; we could, however, in some
future version<br>
of obscore, introduce a multi-valued
"other_dataproduct_types" in the<br>
style of EPN-TAP -- but that's a different
discussion.<br>
<br>
In what I'm drafting for SimpleDALRegExt, SSA
services can be<br>
annotated with zero or more dataproductType elements
(current<br>
internal draft: <a
href="http://docs.g-vo.org/SimpleDALRegExt.pdf"
target="_blank" moz-do-not-send="true">http://docs.g-vo.org/SimpleDALRegExt.pdf</a>
or on<br>
volute). More on this later on the Registry list.<br>
<br>
Thanks,<br>
<br>
Markus</p>
</div>
</div>
</div>
</div>
</blockquote>
</div>
</blockquote>
</body>
</html>