iocean-discuss Mailing List for Inverse Ocean Model
Brought to you by:
ezaron
You can subscribe to this list here.
2007 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(3) |
Nov
|
Dec
|
---|---|---|---|---|---|---|---|---|---|---|---|---|
2008 |
Jan
(1) |
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
From: Ed Z. <za...@ce...> - 2008-01-29 06:19:30
|
The file at IOM-root/software/fortran/iom_handwritten/iom_comp_penalty.F90 which computes the prior value of the penalty function has a significant update. The model forcing error covariance can be declared as either "diagonal" or "full". This routine was allocating space for the full matrix representation, even when a diagonal covariance matrix was used. FIXED. I have added a patch file IOM-root/software/fortran/iom_handwritten/patch.iom_comp_penalty in case you just want to patch your version in the working part of the IOM tree, e.g., IOM-root/Inverse_MODEL/iom/SRC_iom_handwritten/ -Ed |
From: Ed Z. <ez...@co...> - 2007-10-15 22:52:21
|
Julia, How about having the link "Technical Support" under "downloads" point to http://sourceforge.net/projects/iocean/ and place a link on the "reseach" "overview" page, e.g., "New: The IOM is now a hosted project on sourceforge: http:// sourceforge.net/projects/iocean/" - or - If you can fit "Sourceforge Project" in the blue navigation bar, I would put it on all the pages. -Ed ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Edward D. Zaron Research Associate Hydrodynamic Process and Ecosystems Research Group Department of Civil and Environmental Engineering Portland State University Portland, OR 97207-0751 Phone: (503)-725-2435 FAX: (503)-725-5950 za...@ce... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ On Oct 15, 2007, at 3:40 PM, Julia Muccino Cornuelle wrote: > Hi Ed, > > Do you have a preference regarding where on the site to put the > link? I > thought on the home page, and again on the research overview page. > Let me > know if this is consistent with your plan and I'll do it right away. > > Julia > > >> >> Hello Everyone, >> >> I have created a sourceforge project for the iom. The project page is >> located here: >> >> http://sourceforge.net/projects/iocean/ >> >> ------------------------------------------- >> >> The web page and source code are still hosted on our own servers: >> >> 1) The web page for the project, http://iocean.sourceforge.net, is a >> redirect to http://iom.asu.edu. >> >> 2) The source code repository is browseable at http:// >> ocean.cee.pdx.edu/~ezaron/websvn. >> >> The project page at sourceforge provides a mailing list (iocean- >> di...@li...), a bugtracker, and space for a wiki. I >> have CC'd this email to the mailing list, and I have entered Hans >> email regarding diff into the bugtracker. >> >> I think it would be best if we could start using the iocean-discuss >> list for support requests and user reports so that we can avoid >> repetition in individual reports to Boon or Andrew or the IOM >> developers. Likewise, I think it would be best combine bug reports >> with an email to the list and a bugtracker submission. Otherwise >> these things are bound to be lost. >> >> ------------------------------------------- >> >> I would like to email the following to the participants in the last >> IOM workshop: >> >> >> Hello IOMers, >> >> I have created a sourceforge project for the IOM. The project page is >> located here: >> >> http://sourceforge.net/projects/iocean/ >> >> >> You will find that the project has two email lists: >> >> 1) iocean-discuss: For general discussion and support concerning IOM >> installation, use, etc.; and >> >> 2) iocean-announce: For IOM software release announcements. >> >> The benefit of having these lists is that they are archived and >> searchable, so that we can build a base of shared documentation for >> the IOM. You are welcome to join either list. >> >> It would be best if we could start using the iocean-discuss list for >> support requests and user reports. Doing so will avoid repetition in >> individual reports to Boon or Andrew or the IOM developers. Likewise, >> I think it would be best combine bug reports with an email to the >> list and a bugtracker submission. Otherwise these things are bound to >> get lost. >> >> At the project page you will see that the code is now in a version >> control system. Please email me if you would like to contribute code >> or your local changes to the project. In the short term I will be >> integrating in Boon's recent modifications and adding some of my own. >> Major new releases will be announced on iocean-announce. >> >> Happy IOMing, >> >> -Ed >> >> >> ------------------------------------------------ >> >> NEXT STEPS: >> >> Boon: Can you send me a url to the last development version of the >> IOM you worked on? >> >> Ed: I will write some regression tests for one of the IOM toy solvers >> and attempt to merge Boon's code changes into the version-controlled >> code. >> >> Julia: Can you put a link to the sourceforge project page (http:// >> sourceforge.net/projects/iocean) on the iom site? >> >> Jake: Can you try checking out a working copy from the repository? My >> server has an incorrect self-signed certificate, and this has caused >> problems with some Windows clients in the past. If it causes a >> problem, let me know and I will fix the certificate. Also, would you >> please email me a username and password so I can give you write- >> permission to the repository? >> >> Others: Do you mind subscribing to the iocean-discuss list and >> forwarding the usage or bug reports you have received since the IOM >> workshop? Please remove email addresses and hostnames from the >> forwarded text so that they won't be picked up by spam-bots. I am >> submitting the installation notes I originally sent to Boon to the >> list (other notes are in the /trunk/doc/*.muse files of the IOM >> source tree). As this stuff accumulates in the list, I will create a >> FAQ, time permitting. >> >> Regards, >> >> Ed >> >> >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ >> Edward D. Zaron >> Research Associate >> Hydrodynamic Process and Ecosystems Research Group >> Department of Civil and Environmental Engineering >> Portland State University >> Portland, OR 97207-0751 >> Phone: (503)-725-2435 >> FAX: (503)-725-5950 >> za...@ce... >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ >> >> >> >> Begin forwarded message: >>> From: Rich Loft <lo...@uc...> >>> Date: October 6, 2007 9:29:51 AM PDT >>> To: julia muccino <jmu...@wi...>, Ed Zaron >>> <ez...@co...>, Andrew Bennett >>> <be...@co...>, Martin Erwig >>> <er...@ee...> >>> Subject: strategy for path forward >>> >>> All- >>> >>> I have been thinking a bit since last night and have the following >>> strategy in mind for moving forward. >>> >>> Rich >>> >>> Short term Maintenance Issues >>> 1) Ed Zaron has agreed to stand up a sourceforge site, merge Boon's >>> improvements to the system and set up a bug tracking system. >>> 2) Julia will put a hyperlink into the IOM website to this >>> sourceforge site. >>> 3) We will need a point of contact for answering bug reports, I >>> suggest that Ed serve as this until we can get a student in place. >>> 4) We will need a blurb (on the website and email to our users) to >>> tell them what to do when a software problem is encountered. >>> 5) I suggest that we get a mail list of people who are interested >>> in IOMS and broadcast the existence of this sourceforge site. >>> >>> Maintenance/Marketing Through EOT >>> 1) In the short term, we should focus on education and outreach >>> efforts. This has the twofold effect of building a user community >>> and helping >>> attract funds to maintain the system. An idea that may have >>> traction is that we need to combine the subcritical efforts by data >>> assimilation practitioners to >>> reach out to an interdisciplinary community of scientists trying to >>> do data assimilation through tutorials and education activities. We >>> would ask for money >>> to train an interdisciplinary community to use a variety data >>> assimilation and related techniques. This seems to me a good first >>> step moving forward to get >>> knowledge of IOMS out there in the community. >>> >>> 2) To that end, I am offering to contact Doug Nychka, director of >>> the Institute of Math and the Geosciences (IMAGe) >>> that inform him that there is interest in the UCAR community in an >>> interdisciplinary, multi-method data assimilation workshop. >>> This seems to me to be potentially attractive to him because it is: >>> a) a natural hook into their IMAGe Theme of the Year (TOY) >>> b) a natural target for funding by NSF Geo Directorate divisions. >>> c) involves cost sharing potential that will be attractive to IMAGe. >>> >>> 3) I think NCAR would be receptive to hosting such a workshop. >>> >>> 4) Beside the university communities, potential constituencies to >>> approach for such a tutorial/workshop (either as featured systems >>> or attendees) >>> are >>> >>> Assimilation Systems/Methods - >>> Data Assimilation Research Testbed (DART)/ Ensemble Kalman >>> Filtering - http://www.image.ucar.edu/DAReS/DART/ >>> Inverse Ocean Modeling System (IOMS) 4-D Var - http://iom.asu.edu/ >>> >>> Communities Infrastructure Projects interested in Data Assimilation >>> ESMF http://www.esmf.ucar.edu/ >>> CIG Computational Infrastructure for Geodynamics http:// >>> www.geodynamics.org/ >>> >>> Community Models >>> ROMS/IROMS, etc... >>> GEOFest and/or its CIG replacement >>> >>> 5) Raising money for this interdisciplinary DA workshop/tutorial: I >>> believe we can get some money by passing the hat around the various >>> agencies, their program >>> managers and center directors. This would best be facilitated by >>> generating an unsolicited white paper/proposal on the subject, with >>> budget. It would be best if this would be pitched by >>> as many relevant software infrastructure groups as possible. >>> >>> 6) Preparing for the workshop/tutorial could also be a rationale >>> for including support, perhaps through a student assistants and/or >>> part of Jake's time, for maintaining and upgrading the IOMS system >>> and GUI for use by tutorial attendees. >>> >>> Continued IOM Innovation >>> In parallel with this effort, I think we could engage students and >>> commercial entities in improving aspects of the IOMS toolkit. >>> I am thinking, primarily, of improvement in automatic adjoint >>> generation and improvements in the GUI interface. Martin and I both >>> agree that this is >>> not a research project but more of an engineering project, and >>> therefore requires a different approach than was pursued during the >>> ITR project. >>> >>> 1) Seeking funding for a student engineering project with an >>> engineering university or masters student project at OSU with >>> Martin, to improve the auto differentiator and parametric fortran >>> compiler. This is a place where, again, one could imagine cobbling >>> together a variety of outreach funds, visitor program money and >>> summer internship dollars to pay for an effort. >>> >>> 2) Seeking funding from Microsoft to work on autogenerated >>> software. Beyond the autodifferentiator, >>> an autogenerated GUI's that capture specific science workflows, >>> eliminating extraneous features >>> associated with a general interface. >>> >>> >>> Well, I've got all it downloaded from my brain, let's hear what the >>> feedback is. >>> >>> regards, >>> >>> Rich >>> >>> >> >> >> > |
From: Ed Z. <ez...@co...> - 2007-10-15 22:02:43
|
Hello Everyone, I have created a sourceforge project for the iom. The project page is located here: http://sourceforge.net/projects/iocean/ ------------------------------------------- The web page and source code are still hosted on our own servers: 1) The web page for the project, http://iocean.sourceforge.net, is a redirect to http://iom.asu.edu. 2) The source code repository is browseable at http:// ocean.cee.pdx.edu/~ezaron/websvn. The project page at sourceforge provides a mailing list (iocean- di...@li...), a bugtracker, and space for a wiki. I have CC'd this email to the mailing list, and I have entered Hans email regarding diff into the bugtracker. I think it would be best if we could start using the iocean-discuss list for support requests and user reports so that we can avoid repetition in individual reports to Boon or Andrew or the IOM developers. Likewise, I think it would be best combine bug reports with an email to the list and a bugtracker submission. Otherwise these things are bound to be lost. ------------------------------------------- I would like to email the following to the participants in the last IOM workshop: Hello IOMers, I have created a sourceforge project for the IOM. The project page is located here: http://sourceforge.net/projects/iocean/ You will find that the project has two email lists: 1) iocean-discuss: For general discussion and support concerning IOM installation, use, etc.; and 2) iocean-announce: For IOM software release announcements. The benefit of having these lists is that they are archived and searchable, so that we can build a base of shared documentation for the IOM. You are welcome to join either list. It would be best if we could start using the iocean-discuss list for support requests and user reports. Doing so will avoid repetition in individual reports to Boon or Andrew or the IOM developers. Likewise, I think it would be best combine bug reports with an email to the list and a bugtracker submission. Otherwise these things are bound to get lost. At the project page you will see that the code is now in a version control system. Please email me if you would like to contribute code or your local changes to the project. In the short term I will be integrating in Boon's recent modifications and adding some of my own. Major new releases will be announced on iocean-announce. Happy IOMing, -Ed ------------------------------------------------ NEXT STEPS: Boon: Can you send me a url to the last development version of the IOM you worked on? Ed: I will write some regression tests for one of the IOM toy solvers and attempt to merge Boon's code changes into the version-controlled code. Julia: Can you put a link to the sourceforge project page (http:// sourceforge.net/projects/iocean) on the iom site? Jake: Can you try checking out a working copy from the repository? My server has an incorrect self-signed certificate, and this has caused problems with some Windows clients in the past. If it causes a problem, let me know and I will fix the certificate. Also, would you please email me a username and password so I can give you write- permission to the repository? Others: Do you mind subscribing to the iocean-discuss list and forwarding the usage or bug reports you have received since the IOM workshop? Please remove email addresses and hostnames from the forwarded text so that they won't be picked up by spam-bots. I am submitting the installation notes I originally sent to Boon to the list (other notes are in the /trunk/doc/*.muse files of the IOM source tree). As this stuff accumulates in the list, I will create a FAQ, time permitting. Regards, Ed ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Edward D. Zaron Research Associate Hydrodynamic Process and Ecosystems Research Group Department of Civil and Environmental Engineering Portland State University Portland, OR 97207-0751 Phone: (503)-725-2435 FAX: (503)-725-5950 za...@ce... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Begin forwarded message: > From: Rich Loft <lo...@uc...> > Date: October 6, 2007 9:29:51 AM PDT > To: julia muccino <jmu...@wi...>, Ed Zaron > <ez...@co...>, Andrew Bennett > <be...@co...>, Martin Erwig > <er...@ee...> > Subject: strategy for path forward > > All- > > I have been thinking a bit since last night and have the following > strategy in mind for moving forward. > > Rich > > Short term Maintenance Issues > 1) Ed Zaron has agreed to stand up a sourceforge site, merge Boon's > improvements to the system and set up a bug tracking system. > 2) Julia will put a hyperlink into the IOM website to this > sourceforge site. > 3) We will need a point of contact for answering bug reports, I > suggest that Ed serve as this until we can get a student in place. > 4) We will need a blurb (on the website and email to our users) to > tell them what to do when a software problem is encountered. > 5) I suggest that we get a mail list of people who are interested > in IOMS and broadcast the existence of this sourceforge site. > > Maintenance/Marketing Through EOT > 1) In the short term, we should focus on education and outreach > efforts. This has the twofold effect of building a user community > and helping > attract funds to maintain the system. An idea that may have > traction is that we need to combine the subcritical efforts by data > assimilation practitioners to > reach out to an interdisciplinary community of scientists trying to > do data assimilation through tutorials and education activities. We > would ask for money > to train an interdisciplinary community to use a variety data > assimilation and related techniques. This seems to me a good first > step moving forward to get > knowledge of IOMS out there in the community. > > 2) To that end, I am offering to contact Doug Nychka, director of > the Institute of Math and the Geosciences (IMAGe) > that inform him that there is interest in the UCAR community in an > interdisciplinary, multi-method data assimilation workshop. > This seems to me to be potentially attractive to him because it is: > a) a natural hook into their IMAGe Theme of the Year (TOY) > b) a natural target for funding by NSF Geo Directorate divisions. > c) involves cost sharing potential that will be attractive to IMAGe. > > 3) I think NCAR would be receptive to hosting such a workshop. > > 4) Beside the university communities, potential constituencies to > approach for such a tutorial/workshop (either as featured systems > or attendees) > are > > Assimilation Systems/Methods - > Data Assimilation Research Testbed (DART)/ Ensemble Kalman > Filtering - http://www.image.ucar.edu/DAReS/DART/ > Inverse Ocean Modeling System (IOMS) 4-D Var - http://iom.asu.edu/ > > Communities Infrastructure Projects interested in Data Assimilation > ESMF http://www.esmf.ucar.edu/ > CIG Computational Infrastructure for Geodynamics http:// > www.geodynamics.org/ > > Community Models > ROMS/IROMS, etc... > GEOFest and/or its CIG replacement > > 5) Raising money for this interdisciplinary DA workshop/tutorial: I > believe we can get some money by passing the hat around the various > agencies, their program > managers and center directors. This would best be facilitated by > generating an unsolicited white paper/proposal on the subject, with > budget. It would be best if this would be pitched by > as many relevant software infrastructure groups as possible. > > 6) Preparing for the workshop/tutorial could also be a rationale > for including support, perhaps through a student assistants and/or > part of Jake's time, for maintaining and upgrading the IOMS system > and GUI for use by tutorial attendees. > > Continued IOM Innovation > In parallel with this effort, I think we could engage students and > commercial entities in improving aspects of the IOMS toolkit. > I am thinking, primarily, of improvement in automatic adjoint > generation and improvements in the GUI interface. Martin and I both > agree that this is > not a research project but more of an engineering project, and > therefore requires a different approach than was pursued during the > ITR project. > > 1) Seeking funding for a student engineering project with an > engineering university or masters student project at OSU with > Martin, to improve the auto differentiator and parametric fortran > compiler. This is a place where, again, one could imagine cobbling > together a variety of outreach funds, visitor program money and > summer internship dollars to pay for an effort. > > 2) Seeking funding from Microsoft to work on autogenerated > software. Beyond the autodifferentiator, > an autogenerated GUI's that capture specific science workflows, > eliminating extraneous features > associated with a general interface. > > > Well, I've got all it downloaded from my brain, let's hear what the > feedback is. > > regards, > > Rich > > |
From: Ed Z. <za...@ce...> - 2007-10-15 21:52:10
|
Old Notes from EDZ to BC: > Boon, [I started this last week and finished this morning] > > It looks like maybe there is a duplicate pezhat in the tar file on > lightning --- busted my disk quota! For now will try to install on > tako at OSU: > > Notes on building IOM on tako (ppc mac 10.3.9, ibm xlf): > > 1) edit software/Darwin-xlf90.mk for NETCDF location > > 2) install p5-xml-sax and p5-xml-namespacesupport from macports > > 3) install ghc from macports > > 4) install pf from Martin's site > > 5) edit ./install_system to point to pfc directory at /opt/local/bin > > 6) execute ./install_system > This seems to be compiling the same programs repeatedly ... > hmmm ... Maybe it is looping through the models that were installed? > It might be good if the install_system script could pipe stdout to > a log file, and only output stderr and some indication of the > progress to the screen. > > 7) change the path to java in ./iom > Boon had java 1.6, I have 1.4. Seems to be okay. > > 8) execute ./iom > Cool! The GUI runs. Need to figure out how to make java windows > display remotely via X11. > Can select my system and build. > > 9) select "Outputs:Prior" and "Task:Run" > Error, need kernel filename. > > ... > > I will call... > > -Ed > > > On Feb 21, 2007, at 2:48 PM, Boon Chua wrote: > >> >> Hi Ed, >> >> A copy of inverse pezhat can now be download by >> >> ftp ftp.XXXXXXXX.edu >> >> cd /dist/chua/inverse_pez_hat >> >> get iom_rc1.4_pezhat.tar >> >> >> This version was tested on NCAR Linux cluster lightning >> and the test results are in >> >> /gpfs/lightning/bchua/iom_rc1.4_pezhat/Inverse_pezhat/outputs >> >> A lightning version is in /gpfs/lightning/bchua >> (iom_rc1.4_pezhat.tar) >> >> Also >> >> setenv PERL5LIB /contrib/2.6/perlmods/lib/perl5:/contrib/2.6/ >> perlmods/lib >> /perl5/site_perl:/contrib/2.6/perlmods/lib/perl5/site_perl/5.8.3 >> >> >> Please give me a call regarding installation details and etc. >> >> >> Regards, >> >> Boon >> > > |