subqueries no longer working?

Question

I am running into some issues with ADQL queries and would appreciate your assistance.

For one, I get the error 'sub-select not supported in FROM clause'

but I am following the example from this helpdesk response last year:

https://datalab.noirlab.edu/help/index.php?qa=816&qa_1=crossmatching-various-catalogues-various-datalab-tables&show=818#c818

Did subselect work in the past and now no longer works?  Also, I struggled a bit with the lack of support for things like CONTAINS, POINT, CIRCLE, etc.   I understand that you do offer some alternatives like q3c_radial_query but this is not really ADQL then?

As a separate question, are you planning to provide a pre-crossmatched table between ALLWISE and DES_DR2?

Answer 1

Thank you for reaching out. Regarding your questions:

- Sub-selects do work with SQL (i.e. the PostgreSQL syntax) and for sync queries. They are not supported for ADQL or for async queries.

- Correct, our q3c functions are a Postgres extension, and work with SQL only. Our ADQL parser currently does not support functions such as CONTAINS, POINT, or CIRCLE.

- A pre-crossmatched table between AllWISE and DES_DR2 (and the inverse table too) are being generated right now, and should become available in a few days. This was on our to-do anyway, so thank you for the extra push.

Hoping that this is helpful,

Robert for the DL team

Comments

Hi, I have to correct myself: the pre-crossmatched tables being created right now are between AllWISE and NSC_DR2, not DES_DR2. However, we are also currently determining which additional pre-crossmatched tables to generate by default in the future, and DES_DR2 would be very high at the top of the list. So, it will come, but not within "a few days", as I initially indicated. My apologies.

Robert

---

Thank you for your detailed reply, that is very helpful!

There already appears to be a table allwise.x1p5__source__nsc_dr2__object.  Is this the one you are referring to as being created now (maybe it just finished after your last message?) or is this something different?  Do you think joining allwise with nsc_dr2 and then nsc_dr2 with des_dr2 will achieve the same result as directly crossmatching allwise with des_dr2 or will the former be less complete?

I also see nsc_dr2.x1p5__object__ls_dr8__tractor_primary and allwise.x1p5__source__ls_dr8__tractor_primary, but do you know if similar tables for ls_dr9 might be coming soon?