You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
My new PhD student, George Cann (in cc:) started a project 6 weeks ago
sponsored by the UK Space Agency on developing retrieval code for 13CH4
isotopologue retrieval in the Martian atmosphere using data from
instruments onboard the ESA Trace Gas Orbiter.
I previously collaborated with Hartmut Boesch at Leicester using his
code on CEMS for 13CH4 from GOSAT data. Hartmut wrote part of this code
when he was a NRC fellow many years ago at JPL.
For George's PhD project, we want to combine your OCO-2 retrieval code
with the Oxford Reference Forward Model (ORFM) to develop a standalone
13CH4 retrieval system for solar occultation data.
George downloaded the software from the github and has been working on
the hints and tips provided at
However, he has hit a brick-wall after several weeks of trying to
compile and link this software on MacOS. Unfortunately, none of the
developer's notes say what the original target system was so it may be
that this code was developed on RHEL or PC Windows and was never
intended to run on MacOS?
George has access to RHEL blade systems so running on linux might be
much easier. However, I wanted to check first and also to ask if you
would be prepared to answer his key questions so he can start to make
some progress?
The text was updated successfully, but these errors were encountered:
Yes, the software is written and regularly tested on a variety of linux systems. We have built on the MacOS in the past, but it has been a while and isn't officially supported. Thanks for pointing this out, that is a pretty basic thing for us to have left out of the documentation. I filed a git issue for this.
I'd recommend George look at the user guide build directions at https://nasa.github.io/RtRetrievalFrameworkDoc/compilation.html#building, this is the documentation for a user just building the system. The documentation he was looking at is really for developers (i.e., how do you interact with the build system if you are adding new files or something like that). There users guide describes a single script "./nonjpl_build.script" which should work for building the system - this is a good place to start.
Our development system is Centos 6, we also regularly test on Fedora and Ubuntu. I'd expect the software can be built on pretty much any linux system with the required dependencies.
Either I or James McDuffie (CCed on this email) would be happy to answer questions.
How important is the MacOS support? Is this just a nice to have, or something that you particularly need? James has some fraction of burden money to support the retrieval framework, and adding back in and regularly testing for MacOS support might be something to add if there is a need for this - we had just dropped this over time as a low priority at JPL since we generally run on linux.
Email from Peter:
Hi Mike
My new PhD student, George Cann (in cc:) started a project 6 weeks ago
sponsored by the UK Space Agency on developing retrieval code for 13CH4
isotopologue retrieval in the Martian atmosphere using data from
instruments onboard the ESA Trace Gas Orbiter.
I previously collaborated with Hartmut Boesch at Leicester using his
code on CEMS for 13CH4 from GOSAT data. Hartmut wrote part of this code
when he was a NRC fellow many years ago at JPL.
For George's PhD project, we want to combine your OCO-2 retrieval code
with the Oxford Reference Forward Model (ORFM) to develop a standalone
13CH4 retrieval system for solar occultation data.
George downloaded the software from the github and has been working on
the hints and tips provided at
http://nasa.github.io/RtRetrievalFrameworkDoc/developer_information.html
However, he has hit a brick-wall after several weeks of trying to
compile and link this software on MacOS. Unfortunately, none of the
developer's notes say what the original target system was so it may be
that this code was developed on RHEL or PC Windows and was never
intended to run on MacOS?
George has access to RHEL blade systems so running on linux might be
much easier. However, I wanted to check first and also to ask if you
would be prepared to answer his key questions so he can start to make
some progress?
The text was updated successfully, but these errors were encountered: