Dear all,

The deprecation warning is there because we will remove
abs_xsec_per_speciesAddLines before Arts 3.0 is released.  It is just
warning you that some code will have to change in the future.  Ignore it
for the purpose of what you want to do here.

About the speed.  The new code should be comparable and often faster than
the old code (depending on your line-data inputs).

There are two caveats to that though.
  1)  Hitran2020 is much larger than previous copies.
  2)  You will not see parallelism unless you have enough temperature
perturbations.  There is a pending pull-request to fix this.

As Oliver points out, the recommended data to use is the
arts-cat-data/lines/ data.  This is based on Hitran2020 with a few
modifications we found necessary to improve the quality of the catalog.

//Richard

Den mån 14 feb. 2022 kl 13:49 skrev Pengwang Zhai <pwz...@umbc.edu>:

> Hello, Mattia,
>
> Thanks for the update. I have to admit that I did not use the line.xml
> file. I used the H2O file downloaded from HITRAN 2020. Originally I said
> “it seems working”, today I still cannot definitely confirm whether it
> works from my side, as ARTS 2.5 did not finish the calculation after
> several days. Maybe I am trying to calculate too many frequency lines,
> though I believe ARTS 2.3 only took several hours to build the same lookup
> table.
>
> I did see the same warning that you saw regarding
> abs_xsec_per_speciesAddLines.
>
> Yours,
>
> Pengwang
>
> ________________
> Pengwang Zhai (he/his/him)
> Graduate Program Director, Atmospheric Physics,
> Associate Professor, Physics Department, UMBC
> pwz...@umbc.edu
>
>
>
>
>
>
> > On Feb 14, 2022, at 5:00 AM, Mattia Sabatini <
> mattia.sabat...@artov.ismar.cnr.it> wrote:
> >
> > Hello again Pengwang,
> >
> > I tested controlfile TestAbs.arts with your edit:
> >
> > AgendaSet( abs_xsec_agenda ){
> >    abs_xsec_per_speciesInit
> >    abs_xsec_per_speciesAddConts
> >    lbl_checkedCalc
> >    abs_xsec_per_speciesAddLines
> > }
> >
> > And with the following abs species:
> >
> > abs_speciesSet( species=[ "H2O", "H2O-PWR98", "O2-PWR93",
> "N2-SelfContStandardType" ] ).
> >
> > I thought that by adding abs_xsec_per_speciesAddLines in
> abs_xsec_agenda, just like Richard suggested me to do in my post (
> https://www.mail-archive.com/arts_users.mi@lists.uni-hamburg.de/msg00471.html),
> would solve the problem you were experiencing of having a lookup table
> filled with zeros. Unfortunately this is still an issue for me, even with
> the above settings: the H2O cross section values are all zeros, while
> values for "H2O-PWR98", "O2-PWR93", "N2-SelfContStandardType” are not.
> Also, during the execution I had this “Deprecated function warning”:
> >
> > abs_xsec_per_speciesAddLines is deprecated since 2021-07-13
> >
> > This function is no longer up to date.  It only exists to satisfy lookup
> table calculations before these are updated.
> > Once the lookup table calculations are up-to-date, this function is
> fully replaced with propmat_clearskyAddLines, with better functionality
> >
> > Apparently, abs_xsec_per_speciesAddConts is doing its job but
> abs_xsec_per_speciesAddLines does not. Did you had the chance to test it
> too?
> >
> > Mattia
> >
> >
> >
> >
> >
> >
> > Thanks, Oliver and Mattia, for you help.
> >
> > After studying Mattia and Richard's posts, I figured out the following:
> >
> > Comment out line 15 of TestAbs.arts:
> >
> > #Copy(abs_xsec_agenda, abs_xsec_agenda__noCIA)
> >
> > And insert the following to line 16:
> >
> > AgendaSet( abs_xsec_agenda ){
> >    abs_xsec_per_speciesInit
> >    abs_xsec_per_speciesAddConts
> >    lbl_checkedCalc
> >    abs_xsec_per_speciesAddLines
> > }
> >
> > Now it seems working, though I have not yet got the chance to check the
> outputs as it is still running.
> >
> > Yours
> >
> > Pengwang
> >
> >
> >
> >
> >
> >
> >
> > > On Feb 11, 2022, at 2:06 AM, Lemke, Oliver <
> oliver.le...@uni-hamburg.de> wrote:
> > >
> > > Hi Mattia, hi Pengwang,
> > >
> > > Thanks Mattia for helping out. :-)
> > >
> > > Since you were referring to Richard's earlier post, I'll take this
> opportunity to point out the searchable archive of this list, which might
> come in handy at times:
> > >
> > > https://www.mail-archive.com/arts_users.mi@lists.uni-hamburg.de/
> > >
> > > Here is the post Mattia was referring to:
> > >
> > >
> https://www.mail-archive.com/arts_users.mi@lists.uni-hamburg.de/msg00471.html
> > >
> > > Cheers,
> > > Oliver
> > >
> > >
> > >> On 10 Feb 2022, at 17:38, mattia.sabat...@artov.ismar.cnr.it wrote:
> > >>
> > >> Pengwang,
> > >>
> > >> sorry for my partial answer. Recently I compiled ARTS 2.5 and I had a
> similar issue concerning the absorption lookup table. Richard Larsson,
> helped me replying to my mail sent in this mailing list, suggesting to add
> abs_xsec_per_speciesAddLines into my cross-section agenda.
> > >>
> > >> He also added:
> > >>
> > >> "We are currently in a transition in the 2.5-branch of moving away
> from abs_xsec_agenda entirely.  There are some lingering problems when
> using lookup table calculations at this time.  Mainly, since the line
> calculations should happen in propmat_clearsky_agenda now, they are not
> part of any of the default cross-section agendas as in the past.  The
> lookup generation interface is currently under active development, so this
> interface might change in the near future."
> > >>
> > >> Best regards,
> > >> Mattia
> > >>
> > >>
> > >>
> > >>
> > >>
> > >> Pengwang Zhai <pwz...@umbc.edu> ha scritto:
> > >>
> > >>> Thanks, Mattia. Would you advise how TestAbs.arts can be revised to
> include line-by-line calculation?
> > >>>
> > >>> Note that the behavior of TestAbs.arts is different in arts2.3,
> which does calculate line-by-line absorption coefficients.
> > >>>
> > >>> I do not want continua, as I am mainly interested in the visible
> spectra, which is out of most of those continuum models. Based on my
> experience with previous arts version, arts does not check the spectrum
> limits for the continuum models. I am not sure whether arts 2.5 has been
> improved on this.
> > >>>
> > >>> Pengwang
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>> On Feb 10, 2022, at 9:46 AM, mattia.sabat...@artov.ismar.cnr.it
> wrote:
> > >>>>
> > >>>> Hi Pengwang,
> > >>>>
> > >>>> the controlfile TestAbs.arts, in line 15, has the following:
> > >>>> Copy(abs_xsec_agenda, abs_xsec_agenda__noCIA)
> > >>>>
> > >>>> I checked in controlfiles/general/agendas.arts for what
> abs_xsec_agenda__noCIA does:
> > >>>> AgendaCreate( abs_xsec_agenda__noCIA )
> > >>>> AgendaSet( abs_xsec_agenda__noCIA ){
> > >>>> abs_xsec_per_speciesInit
> > >>>> abs_xsec_per_speciesAddConts
> > >>>> }
> > >>>>
> > >>>> Therefore the agenda calculates absorption for continua tag only
> (see
> https://atmtools.github.io/arts-docs-master/docserver/methods/abs_xsec_per_speciesAddConts.html).
> You selected H2O, and by doing this my guess is that you are not
> considering continuum, as it is written in lines 36-37 of TestAbs.arts.
> > >>>>
> > >>>> I hope that this will help you,
> > >>>>
> > >>>> Mattia
> > >>>>
> > >>>>
> > >>>> Pengwang Zhai <pwz...@umbc.edu> ha scritto:
> > >>>>
> > >>>>> Hello, ARTS community,
> > >>>>>
> > >>>>> I downloaded and compiled the latest version of arts. Now I tested
> the the creation of abs_lookup with the example control file located in:
> > >>>>>
> > >>>>> arts/controlfiles/artscomponents/absorption/TestAbs.arts
> > >>>>>
> > >>>>> I only modified two occurrences of:
> > >>>>>
> > >>>>> abs_speciesSet( species=[ "H2O-PWR98",
> > >>>>>                        "O2-PWR93",
> > >>>>>                        "N2-SelfContStandardType" ] )
> > >>>>>
> > >>>>> to:
> > >>>>>
> > >>>>> abs_speciesSet( species=[ "H2O" ] )
> > >>>>>
> > >>>>> and run
> > >>>>>
> > >>>>> arts TestAbs.arts
> > >>>>>
> > >>>>> The resultant abs_lookup are all ZEROs.
> > >>>>>
> > >>>>> Any help?
> > >>>>>
> > >>>>> More background information: I used arts 2.3 to create abs_lookup
> for H2O, CO2, etc. in the visible by reading from HITRAN 2012. The baseline
> example was TestAbs.arts provided by the arts installation. Now I need to
> revisit the calculation based on HITRAN 2020, and the lookup table
> calculation seems not working with arts 2.5. I greatly appreciate it if you
> could provide a working example on how to create a abs_lookup with arts 2.5.
> > >>>>>
> > >>>>> Pengwang
> > >
>
> _______________________________________________
> arts_users.mi mailing list
> arts_users.mi@lists.uni-hamburg.de
> https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi
>
_______________________________________________
arts_users.mi mailing list
arts_users.mi@lists.uni-hamburg.de
https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi

Reply via email to