Re: [arts-users] Temperature Retrievals of Oxygen emission lines with Zeeman effect and line mixing
Hi Witali, Great that it worked with GCC. The problem with Clang is a known issue on operating systems that link by default against the GNU libstdc++. In the case of Ubuntu, compiling with clang requires explicit linking against LLVM's libc++ by setting `export CXXFLAGS=-stdlib=libc++` before calling cmake. I'm working on a pull request[1] that updates our cmake configuration to output proper errors if the available compiler is too old and will also add the automatic switch to libc++ if clang is used. Thanks for your feedback and good luck with the line mixing! Cheers, Oliver [1] https://github.com/atmtools/arts/pull/526 > On 14 Sep 2022, at 17:48, witali.kroc...@unibe.ch wrote: > > I was running Ubuntu 20.04LTS which only offers gcc 9 and clang 10. Upgrading > gcc and clang on 20.04 was not possible (at least not easily). After > upgrading to 22.10 which does provide gcc 11 and clang 14 I managed to > install ARTS, but only using gcc. When trying to use clang as directed in > "LLVM/Clang compiler" section on the gitpage there was an error: > "fatal error: instantiating fold expression with 2003 arguments exceeded > expression nesting limit of 256" > > Will try to run the line mixing module know. > > Best regards > Witali smime.p7s Description: S/MIME cryptographic signature
Re: [arts-users] Temperature Retrievals of Oxygen emission lines with Zeeman effect and line mixing
I was running Ubuntu 20.04LTS which only offers gcc 9 and clang 10. Upgrading gcc and clang on 20.04 was not possible (at least not easily). After upgrading to 22.10 which does provide gcc 11 and clang 14 I managed to install ARTS, but only using gcc. When trying to use clang as directed in "LLVM/Clang compiler" section on the gitpage there was an error: "fatal error: instantiating fold expression with 2003 arguments exceeded expression nesting limit of 256" Will try to run the line mixing module know. Best regards Witali Von: Lemke, Oliver Gesendet: Mittwoch, 14. September 2022 14:31:51 An: Krochin, Witali (IAP) Cc: ARTS Users List Betreff: Re: [arts-users] Temperature Retrievals of Oxygen emission lines with Zeeman effect and line mixing Hi Witali, That is indeed strange. Which compiler are you using? You need at least GCC 11 or LLVM/Clang 13. Cheers, Oliver > On 14 Sep 2022, at 14:22, > wrote: > > Thank you very much for the support. > > After installing the master branch and running the 'make check' command, I > got a weird error message that read: > > [ 10%] Building CXX object > src/CMakeFiles/make_workspace_memory_handler_cc.dir/make_workspace_memory_handler_cc.cc.o > In file included from /opt/arts-master/src/matpackI.h:102, > from /opt/arts-master/src/methods.h:32, > from /opt/arts-master/src/global_data.h:31, > from > opt/arts-master/src/make_workspace_memory_handler_cc.cc:24: > /opt/arts-master/src/matpack_concepts.h:4:10: fatal error: concepts: No such > file or directory > 4 | #include > | ^~ > compilation terminated. > > I'm sure this file exists and my machine should meet the requirements as well. > Reinstalling ARTS 2.4 worked fine before. > > > Is this a known issue? > > Best regards > Witali
Re: [arts-users] Temperature Retrievals of Oxygen emission lines with Zeeman effect and line mixing
Hi Witali, That is indeed strange. Which compiler are you using? You need at least GCC 11 or LLVM/Clang 13. Cheers, Oliver > On 14 Sep 2022, at 14:22, > wrote: > > Thank you very much for the support. > > After installing the master branch and running the 'make check' command, I > got a weird error message that read: > > [ 10%] Building CXX object > src/CMakeFiles/make_workspace_memory_handler_cc.dir/make_workspace_memory_handler_cc.cc.o > In file included from /opt/arts-master/src/matpackI.h:102, > from /opt/arts-master/src/methods.h:32, > from /opt/arts-master/src/global_data.h:31, > from > opt/arts-master/src/make_workspace_memory_handler_cc.cc:24: > /opt/arts-master/src/matpack_concepts.h:4:10: fatal error: concepts: No such > file or directory > 4 | #include > | ^~ > compilation terminated. > > I'm sure this file exists and my machine should meet the requirements as well. > Reinstalling ARTS 2.4 worked fine before. > > > Is this a known issue? > > Best regards > Witali smime.p7s Description: S/MIME cryptographic signature
Re: [arts-users] Temperature Retrievals of Oxygen emission lines with Zeeman effect and line mixing
Thank you very much for the support. After installing the master branch and running the 'make check' command, I got a weird error message that read: [ 10%] Building CXX object src/CMakeFiles/make_workspace_memory_handler_cc.dir/make_workspace_memory_handler_cc.cc.o In file included from /opt/arts-master/src/matpackI.h:102, from /opt/arts-master/src/methods.h:32, from /opt/arts-master/src/global_data.h:31, from opt/arts-master/src/make_workspace_memory_handler_cc.cc:24: /opt/arts-master/src/matpack_concepts.h:4:10: fatal error: concepts: No such file or directory 4 | #include | ^~ compilation terminated. I'm sure this file exists and my machine should meet the requirements as well. Reinstalling ARTS 2.4 worked fine before. Is this a known issue? Best regards Witali Von: Lemke, Oliver Gesendet: Mittwoch, 14. September 2022 08:54:20 An: Krochin, Witali (IAP) Cc: Buehler, Stefan; Richard Larsson; ARTS Users List Betreff: Re: [arts-users] Temperature Retrievals of Oxygen emission lines with Zeeman effect and line mixing Hi Witali, The error you're seeing indicates that you're using an older ARTS version. To read the newest version of the arts-cat-data, you also need to compile the latest ARTS development version from the master branch at https://github.com/atmtools/arts . Cheers, Oliver > On 13 Sep 2022, at 17:18, witali.kroc...@unibe.ch wrote: > > Thank you for the quick reply and comments on the updated papers. > > > We will definitely add Mr. Larsson to the list of co-authors when the > research using this dataset is published. > > > Unfortunately, I am already having problems reading the dataset. > I have downloaded the file "O2-66.xml" from > https://arts.mi.uni-hamburg.de/svn/rt/arts-cat-data/trunk/lines/, > but I cannot read the file with any reading function that are known to me. > > > For example, the following line: > "ReadXML(abs_lines, "O2-66.xml")" > > > Generates the following error message: > "Species "" is not a valid species." > > I've tried several other read functions, but I have a feeling I'm missing > something. > > Best regards > Witali
Re: [arts-users] Temperature Retrievals of Oxygen emission lines with Zeeman effect and line mixing
Hi Witali, The error you're seeing indicates that you're using an older ARTS version. To read the newest version of the arts-cat-data, you also need to compile the latest ARTS development version from the master branch at https://github.com/atmtools/arts . Cheers, Oliver > On 13 Sep 2022, at 17:18, witali.kroc...@unibe.ch wrote: > > Thank you for the quick reply and comments on the updated papers. > > > We will definitely add Mr. Larsson to the list of co-authors when the > research using this dataset is published. > > > Unfortunately, I am already having problems reading the dataset. > I have downloaded the file "O2-66.xml" from > https://arts.mi.uni-hamburg.de/svn/rt/arts-cat-data/trunk/lines/, > but I cannot read the file with any reading function that are known to me. > > > For example, the following line: > "ReadXML(abs_lines, "O2-66.xml")" > > > Generates the following error message: > "Species "" is not a valid species." > > I've tried several other read functions, but I have a feeling I'm missing > something. > > Best regards > Witali
Re: [arts-users] Temperature Retrievals of Oxygen emission lines with Zeeman effect and line mixing
Thank you for the quick reply and comments on the updated papers. We will definitely add Mr. Larsson to the list of co-authors when the research using this dataset is published. Unfortunately, I am already having problems reading the dataset. I have downloaded the file "O2-66.xml" from https://arts.mi.uni-hamburg.de/svn/rt/arts-cat-data/trunk/lines/, but I cannot read the file with any reading function that are known to me. For example, the following line: "ReadXML(abs_lines, "O2-66.xml")" Generates the following error message: "Species "" is not a valid species." I've tried several other read functions, but I have a feeling I'm missing something. Best regards Witali Von: arts_users.mi im Auftrag von Buehler, Stefan Gesendet: Montag, 12. September 2022 14:21:05 An: Richard Larsson Cc: arts users mi; Krochin, Witali (IAP) Betreff: Re: [arts-users] Temperature Retrievals of Oxygen emission lines with Zeeman effect and line mixing Dear Witali, since these are unpublished data, I would ask you to include Richard as a coauthor if you intend to publish anything based on this. Best wishes Stefan Am 12.09.2022 um 14:09 schrieb Richard Larsson : Dear Witali, You can use my calculations for line mixing. They are still not published but available online in the arts-cat-data. They have the disadvantage that they are not published but the advantage that they are self-consistent. You can download this data at https://www.radiativetransfer.org/tools/, under the Arts Catalog Data header. The format is simply x0 x1 x2 x3 from table 2.5 of the theory guide at the correct positions by what the xml-tags say. Note that the paper you found has been updated several times. Two updates are at these dois: http://dx.doi.org/10.1016/j.jqsrt.2013.02.019 and https://doi.org/10.1016/j.jqsrt.2019.106798 Note especially that these correct their values in the first paper because they found some theoretical errors. Also note that my implementation in Arts today differs from the one written in these papers because there are some weird typos in there. With hope, --Richard Den mån 12 sep. 2022 kl 13:30 skrev mailto:witali.kroc...@unibe.ch>>: Dear ARTS Team I am currently using the Zeeman module to retrieve stratospheric temperature profiles from oxygen emission lines from fine structure transitions at 52.542 and 53.067 GHz. I correct for the tropospheric influence using a simple tropospheric correction. However, I intend to improve the inversion algorithm by including line-mixing calculations instead. First, I am having trouble figuring out how to get line-mixing data. I found "Makarov, Tretyakov, and Rosenkranz, 2011" as a standard reference for line-mixing data in this frequency range. However, it seems to me that only the emission complex at 60 GHz is covered there. Is there any line mixing data for the emission lines at 52.542 and 53.067 GHz? Also, I would need some guidance on the implementation of the line mixing module. I guess the data set should be in a certain format, which I did not find in the ARTS manual or documentation (maybe I missed something?). With kind regards Witali Krochin
Re: [arts-users] Temperature Retrievals of Oxygen emission lines with Zeeman effect and line mixing
Dear Witali, since these are unpublished data, I would ask you to include Richard as a coauthor if you intend to publish anything based on this. Best wishes Stefan > Am 12.09.2022 um 14:09 schrieb Richard Larsson : > > > Dear Witali, > > You can use my calculations for line mixing. They are still not published > but available online in the arts-cat-data. They have the disadvantage that > they are not published but the advantage that they are self-consistent. You > can download this data at https://www.radiativetransfer.org/tools/, under the > Arts Catalog Data header. > > The format is simply x0 x1 x2 x3 from table 2.5 of the theory guide at the > correct positions by what the xml-tags say. > > Note that the paper you found has been updated several times. Two updates > are at these dois: http://dx.doi.org/10.1016/j.jqsrt.2013.02.019 and > https://doi.org/10.1016/j.jqsrt.2019.106798 > > Note especially that these correct their values in the first paper because > they found some theoretical errors. Also note that my implementation in Arts > today differs from the one written in these papers because there are some > weird typos in there. > > With hope, > --Richard > > Den mån 12 sep. 2022 kl 13:30 skrev : >> Dear ARTS Team >> >> >> >> I am currently using the Zeeman module to retrieve stratospheric temperature >> profiles from oxygen emission lines from fine structure transitions at >> 52.542 and 53.067 GHz. I correct for the tropospheric influence using a >> simple tropospheric correction. However, I intend to improve the inversion >> algorithm by including line-mixing calculations instead. >> >> First, I am having trouble figuring out how to get line-mixing data. I found >> "Makarov, Tretyakov, and Rosenkranz, 2011" as a standard reference for >> line-mixing data in this frequency range. However, it seems to me that only >> the emission complex at 60 GHz is covered there. Is there any line mixing >> data for the emission lines at 52.542 and 53.067 GHz? >> >> >> >> Also, I would need some guidance on the implementation of the line mixing >> module. I guess the data set should be in a certain format, which I did not >> find in the ARTS manual or documentation (maybe I missed something?). >> >> >> >> With kind regards >> >> Witali Krochin >> >> smime.p7s Description: S/MIME cryptographic signature
Re: [arts-users] Temperature Retrievals of Oxygen emission lines with Zeeman effect and line mixing
Dear Witali, You can use my calculations for line mixing. They are still not published but available online in the arts-cat-data. They have the disadvantage that they are not published but the advantage that they are self-consistent. You can download this data at https://www.radiativetransfer.org/tools/, under the Arts Catalog Data header. The format is simply x0 x1 x2 x3 from table 2.5 of the theory guide at the correct positions by what the xml-tags say. Note that the paper you found has been updated several times. Two updates are at these dois: http://dx.doi.org/10.1016/j.jqsrt.2013.02.019 and https://doi.org/10.1016/j.jqsrt.2019.106798 Note especially that these correct their values in the first paper because they found some theoretical errors. Also note that my implementation in Arts today differs from the one written in these papers because there are some weird typos in there. With hope, --Richard Den mån 12 sep. 2022 kl 13:30 skrev : > Dear ARTS Team > > > I am currently using the Zeeman module to retrieve stratospheric > temperature profiles from oxygen emission lines from fine structure > transitions at 52.542 and 53.067 GHz. I correct for the tropospheric > influence using a simple tropospheric correction. However, I intend to > improve the inversion algorithm by including line-mixing calculations > instead. > > First, I am having trouble figuring out how to get line-mixing data. I > found "Makarov, Tretyakov, and Rosenkranz, 2011" as a standard reference > for line-mixing data in this frequency range. However, it seems to me that > only the emission complex at 60 GHz is covered there. Is there any line > mixing data for the emission lines at 52.542 and 53.067 GHz? > > > > Also, I would need some guidance on the implementation of the line mixing > module. I guess the data set should be in a certain format, which I did not > find in the ARTS manual or documentation (maybe I missed something?). > > > > With kind regards > > Witali Krochin > >
