From luigi at lambrate.inaf.it Fri Mar 4 08:42:18 2011 From: luigi at lambrate.inaf.it (Luigi Paioro) Date: Fri, 04 Mar 2011 14:42:18 +0100 Subject: [AstroPy] SAMPy 1.2.1 Message-ID: <4D70EC3A.4060106@lambrate.inaf.it> Dear all, with this message I announce the release of SAMPy 1.2.1 (supporting the experimental Web Profile [*]) and some news related with it. First of all, SAMPy is now directly distributed through PyPI Package Index service: http://pypi.python.org/pypi/sampy/ where the sources can be downloaded in tar.gz format or zip format. It can also be installed using pip installation tool for Python packages, available here: http://pip.openplans.org/ With pip, SAMPy can be easily installed using the command: shell > pip install sampy This way, pip tool will search for the latest SAMPy release stored into PyPI repository, download and directly install it. The latest SAMPy API documentation can be reached on-line at this URL: http://packages.python.org/sampy/ It follows the complete set of changes made in SAMPy 1.2.1, taken from the ChangeLog: - Updated version supporting the Web Profile: a) -w/--web-profile option added to sampy.py script as stand-alone application and b) web_profile optional argument added to SAMPHubServer class, in order to run the Hub enabling the Web Profile. - Bug fixed. Host name better resolved on proxied networks. - `sampy` script added: it is just a simple script importing SAMPy module and executing it calling the main routine. `sampy` has been thought to be a replacement of sampy.py as stand alone program (still running though). - README and setup.py files updated. - SAMPy icons added. - Utility scripts and examples moved in more appropriate directories. - SAMPy Hub metadata modified. - Other minor changes and improvements. For any comment, bug report, etc. please contact me. With my Best Regards, Luigi Paioro [*] http://www.star.bristol.ac.uk/~mbt/websamp/ -- Luigi Paioro INAF - IASF Milano Via Bassini 15, I-20133 Milano, Italy Phone (+39) 02 23 699 470 Fax (+39) 02 26 660 17 Site http://cosmos.iasf-milano.inaf.it/luigi/ From miguel.deval at gmail.com Mon Mar 7 11:02:37 2011 From: miguel.deval at gmail.com (Miguel de Val-Borro) Date: Mon, 7 Mar 2011 17:02:37 +0100 Subject: [AstroPy] PyCon Message-ID: <20110307160237.GE2339@poincare> Hi, I will present a poster at PyCon about Herschel data reduction and visualization using python tools. I have put a draft of the poster here: http://www.mps.mpg.de/data/outgoing/deval/poster.pdf If you will be at PyCon please stop by during the poster session. I will be in Atlanta from Wednesday to Monday afternoon. Regards, Miguel From astropy at liska.ath.cx Mon Mar 14 11:52:46 2011 From: astropy at liska.ath.cx (Ole Streicher) Date: Mon, 14 Mar 2011 16:52:46 +0100 Subject: [AstroPy] Ubuntu packages and state of AstroLib Message-ID: <4D7E39CE.907@liska.ath.cx> Dear list, are there Ubuntu packages available for AstroLib? And/or are the packages for Debian or are there any efforts to create such? Since I need pywcs for Ubuntu, I would volunteer to create such a package, but I'd like to avoid doubling effort. In case I would need to create them myself: it seems a bit unclear on how to get started. The astrolib homepage does not show any releases; however there seem to be some openSUSE package built from a (released??) version http://stsdas.stsci.edu/astrolib/pywcs-1.9-4.4.4.tar.gz What is the current state of this software? Cheers Ole From jturner at gemini.edu Mon Mar 14 12:52:03 2011 From: jturner at gemini.edu (James Turner) Date: Mon, 14 Mar 2011 12:52:03 -0400 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <4D7E39CE.907@liska.ath.cx> References: <4D7E39CE.907@liska.ath.cx> Message-ID: <4D7E47B3.1080902@gemini.edu> > are there Ubuntu packages available for AstroLib? And/or are the packages > for Debian or are there any efforts to create such? I don't think so so far, but... > Since I need pywcs for Ubuntu, I would volunteer to create such a > package, but I'd like to avoid doubling effort. Be aware that pywcs is also being distributed with STScI Python now, so you might even have it already. Of course it would be good to have Astrolib easily installable, rather than just sitting in the repository for developers to grab. In future I expect it will be part of the easy-install Python distribution we have started working on. Unfortunately it has taken rather a long time to get the documentation server set up that was requested last year, to help make Astrolib more accessible. STScI are more familiar with what's already there than I am and might have better guidance re. their intentions etc. (they are probably reading this). Cheers, James. > In case I would need to create them myself: it seems a bit unclear on > how to get started. The astrolib homepage does > not show any releases; however there seem to be some openSUSE package > built from a (released??) version > http://stsdas.stsci.edu/astrolib/pywcs-1.9-4.4.4.tar.gz > > What is the current state of this software? > > Cheers > > Ole > _______________________________________________ > AstroPy mailing list > AstroPy at scipy.org > http://mail.scipy.org/mailman/listinfo/astropy From jturner at gemini.edu Mon Mar 14 12:52:09 2011 From: jturner at gemini.edu (James Turner) Date: Mon, 14 Mar 2011 12:52:09 -0400 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <4D7E39CE.907@liska.ath.cx> References: <4D7E39CE.907@liska.ath.cx> Message-ID: <4D7E47B9.3000304@gemini.edu> > are there Ubuntu packages available for AstroLib? And/or are the packages > for Debian or are there any efforts to create such? I don't think so so far, but... > Since I need pywcs for Ubuntu, I would volunteer to create such a > package, but I'd like to avoid doubling effort. Be aware that pywcs is also being distributed with STScI Python now, so you might even have it already. Of course it would be good to have Astrolib easily installable, rather than just sitting in the repository for developers to grab. In future I expect it will be part of the easy-install Python distribution we have started working on. Unfortunately it has taken rather a long time to get the documentation server set up that was requested last year, to help make Astrolib more accessible. STScI are more familiar with what's already there than I am and might have better guidance re. their intentions etc. (they are probably reading this). Cheers, James. > In case I would need to create them myself: it seems a bit unclear on > how to get started. The astrolib homepage does > not show any releases; however there seem to be some openSUSE package > built from a (released??) version > http://stsdas.stsci.edu/astrolib/pywcs-1.9-4.4.4.tar.gz > > What is the current state of this software? > > Cheers > > Ole > _______________________________________________ > AstroPy mailing list > AstroPy at scipy.org > http://mail.scipy.org/mailman/listinfo/astropy From perry at stsci.edu Mon Mar 14 14:30:38 2011 From: perry at stsci.edu (Perry Greenfield) Date: Mon, 14 Mar 2011 14:30:38 -0400 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <4D7E47B9.3000304@gemini.edu> References: <4D7E39CE.907@liska.ath.cx> <4D7E47B9.3000304@gemini.edu> Message-ID: On Mar 14, 2011, at 12:52 PM, James Turner wrote: >> are there Ubuntu packages available for AstroLib? And/or are the >> packages >> for Debian or are there any efforts to create such? > > I don't think so so far, but... > >> Since I need pywcs for Ubuntu, I would volunteer to create such a >> package, but I'd like to avoid doubling effort. > > Be aware that pywcs is also being distributed with STScI Python now, > so you might even have it already. > > Of course it would be good to have Astrolib easily installable, > rather than just sitting in the repository for developers to grab. > In future I expect it will be part of the easy-install Python > distribution we have started working on. Unfortunately it has taken > rather a long time to get the documentation server set up that was > requested last year, to help make Astrolib more accessible. > > STScI are more familiar with what's already there than I am and > might have better guidance re. their intentions etc. (they are > probably reading this). Actually, I think that the documentation server is working, but other events have kept us from doing much in this area lately. As far as defining releases, I'll need to talk to Mark Sienkiewicz about what that would entail. I'll try to post on that in a day or two. Perry From erik.tollerud at gmail.com Mon Mar 14 14:30:34 2011 From: erik.tollerud at gmail.com (Erik Tollerud) Date: Mon, 14 Mar 2011 11:30:34 -0700 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <4D7E47B9.3000304@gemini.edu> References: <4D7E39CE.907@liska.ath.cx> <4D7E47B9.3000304@gemini.edu> Message-ID: It seems to me that it makes sense to have pywcs available as a separate package on PyPI and thus have it easily packagable on the various distros (and via pip or easy_install). Just like how pyfits is used sometimes as a standalone and sometimes in conjunction with other stsci_python pacakges, there are a lot of pywcs use cases that don't require anything else other than pyfits. I know I've been kept from using it in some places because it's much harder to automate installation of a package if it isn't on PyPI. Maybe this isn't possible due to some kind of dependency I'm not aware of, though. On Mon, Mar 14, 2011 at 9:52 AM, James Turner wrote: >> are there Ubuntu packages available for AstroLib? And/or are the packages >> for Debian or are there any efforts to create such? > > I don't think so so far, but... > >> Since I need pywcs for Ubuntu, I would volunteer to create such a >> package, but I'd like to avoid doubling effort. > > Be aware that pywcs is also being distributed with STScI Python now, > so you might even have it already. > > Of course it would be good to have Astrolib easily installable, > rather than just sitting in the repository for developers to grab. > In future I expect it will be part of the easy-install Python > distribution we have started working on. Unfortunately it has taken > rather a long time to get the documentation server set up that was > requested last year, to help make Astrolib more accessible. > > STScI are more familiar with what's already there than I am and > might have better guidance re. their intentions etc. (they are > probably reading this). > > Cheers, > > James. > > >> In case I would need to create them myself: it seems a bit unclear on >> how to get started. The astrolib homepage ?does >> not show any releases; however there seem to be some openSUSE package >> built from a (released??) version >> http://stsdas.stsci.edu/astrolib/pywcs-1.9-4.4.4.tar.gz >> >> What is the current state of this software? >> >> Cheers >> >> Ole >> _______________________________________________ >> AstroPy mailing list >> AstroPy at scipy.org >> http://mail.scipy.org/mailman/listinfo/astropy > _______________________________________________ > AstroPy mailing list > AstroPy at scipy.org > http://mail.scipy.org/mailman/listinfo/astropy > -- Erik Tollerud From sergiopr at fis.ucm.es Mon Mar 14 18:19:09 2011 From: sergiopr at fis.ucm.es (Sergio Pascual) Date: Mon, 14 Mar 2011 23:19:09 +0100 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <4D7E39CE.907@liska.ath.cx> References: <4D7E39CE.907@liska.ath.cx> Message-ID: Hi, I have created a pywcs package for Fedora, it is still awaiting review (if there are Fedora reviewers in the list, bug number is 654583). I have taken the source code from stsdas.stsci.edu, I have added a patch to use the system version of wcslib instead of the bundled one. There are some other astronomy python packages in Fedora. pyfits and numdisplay are already included. I have packaged ATpy and asciitable also. Hope that helps 2011/3/14 Ole Streicher : > Dear list, > > are there Ubuntu packages available for AstroLib? And/or are the packages > for Debian or are there any efforts to create such? > > Since I need pywcs for Ubuntu, I would volunteer to create such a > package, but I'd like to avoid doubling effort. > > In case I would need to create them myself: it seems a bit unclear on > how to get started. The astrolib homepage does > not show any releases; however there seem to be some openSUSE package > built from a (released??) version > http://stsdas.stsci.edu/astrolib/pywcs-1.9-4.4.4.tar.gz > > What is the current state of this software? > > Cheers > > Ole > _______________________________________________ > AstroPy mailing list > AstroPy at scipy.org > http://mail.scipy.org/mailman/listinfo/astropy > -- Sergio Pascual ? ? ? ? ? ? ? ? ? ? ? ?http://guaix.fis.ucm.es/~spr gpg fingerprint: 5203 B42D 86A0 5649 410A F4AC A35F D465 F263 BCCC Departamento de Astrof?sica -- Universidad Complutense de Madrid (Spain) From astropy at liska.ath.cx Tue Mar 15 09:08:34 2011 From: astropy at liska.ath.cx (Ole Streicher) Date: Tue, 15 Mar 2011 14:08:34 +0100 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: References: <4D7E39CE.907@liska.ath.cx> Message-ID: <4D7F64D2.1090700@liska.ath.cx> Dear Sergio, Am 14.03.2011 23:19, schrieb Sergio Pascual: > I have created a pywcs package for Fedora, it is still awaiting review > (if there are Fedora reviewers in the list, bug number is 654583). This is the version I found on the openSUSE build server. > I have taken the source code from stsdas.stsci.edu, I have added a > patch to use the system version of wcslib instead of the bundled one. Since I already packaged libwcs for Ubuntu, I will probably do the same and use your patch, too. I still dont get the relationship between scipy, astropy and stsci with respect to pywcs. It is developed by STScI as part of their Python package, right? Is the STSci Python project the same as the Astropy project on scipy? Cheers Ole From astropy at liska.ath.cx Thu Mar 17 09:49:14 2011 From: astropy at liska.ath.cx (Ole Streicher) Date: Thu, 17 Mar 2011 14:49:14 +0100 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <4D7E39CE.907@liska.ath.cx> References: <4D7E39CE.907@liska.ath.cx> Message-ID: <4D82115A.9070409@liska.ath.cx> Dear list, I could successfully build packages of pywcs for Ubuntu. They are in my repository (read on there on how to add it to your system): https://launchpad.net/~olebole/+archive/astro Because of the naming conventions of Ubuntu, the package is renamed to "python-pywcs". The source (version 1.9-4.4.4) is taken from stsci, as in the Fedora package made by Sergio Pascual. BTW, this version isn't tagged in the astrolib SVN repository, is it? I also used Sergio's patch to link to the existing libwcs instead of using the builtin one. Since the current libwcs (version 4.7) is used, I needed to remove the underscore in wcsset_() to avoid linkage problems. The repository also contains an Ubuntu package of the libwcs. Packages are available in 32 bit (i386) and 64 bit (amd64) for the 10.04 LTS version (Lucid Lynx) as well as for the current 10.10 version (Maverick Meercat). I will create packages for the upcoming 11.04 (Natty Narwhal) when they reach the Beta testing phase. The main problem for me here is testing: I found only some test code in the astropy subversion repository of pywcs, without documentation, and probably using some features that are not available in pywcs-1.9: UnitConverter, and find_all_wcs(). Also, the test code seem to be not use Pythons "unittest" package? Comments and bug reports on the builds are very welcome. Cheers Ole From mdroe at stsci.edu Thu Mar 17 10:00:57 2011 From: mdroe at stsci.edu (Michael Droettboom) Date: Thu, 17 Mar 2011 14:00:57 +0000 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <4D82115A.9070409@liska.ath.cx> References: <4D7E39CE.907@liska.ath.cx>,<4D82115A.9070409@liska.ath.cx> Message-ID: <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu> The tests are designed for use with nosetests. I admit the tests are rather incomplete -- at STScI pywcs is mainly tested within the context of other tools (e.g. betadrizzle) that use it. It's probably worth investing a little time to add more low-level unit tests. The wcsset_ issue is to support compilation on MS-Windows. As wcslib did not accept the patch upstream, we have had to maintain it in our local copy. But that could probably be improved using #defines or some such to make this easier for packagers to use a system copy of wcslib. BTW -- be careful interchanging the terms "wcslib" and "libwcs" -- they are different projects with different licenses. Cheers, Mike ________________________________________ From: astropy-bounces at scipy.org [astropy-bounces at scipy.org] on behalf of Ole Streicher [astropy at liska.ath.cx] Sent: Thursday, March 17, 2011 9:49 AM To: astropy at scipy.org Subject: Re: [AstroPy] Ubuntu packages and state of AstroLib Dear list, I could successfully build packages of pywcs for Ubuntu. They are in my repository (read on there on how to add it to your system): https://launchpad.net/~olebole/+archive/astro Because of the naming conventions of Ubuntu, the package is renamed to "python-pywcs". The source (version 1.9-4.4.4) is taken from stsci, as in the Fedora package made by Sergio Pascual. BTW, this version isn't tagged in the astrolib SVN repository, is it? I also used Sergio's patch to link to the existing libwcs instead of using the builtin one. Since the current libwcs (version 4.7) is used, I needed to remove the underscore in wcsset_() to avoid linkage problems. The repository also contains an Ubuntu package of the libwcs. Packages are available in 32 bit (i386) and 64 bit (amd64) for the 10.04 LTS version (Lucid Lynx) as well as for the current 10.10 version (Maverick Meercat). I will create packages for the upcoming 11.04 (Natty Narwhal) when they reach the Beta testing phase. The main problem for me here is testing: I found only some test code in the astropy subversion repository of pywcs, without documentation, and probably using some features that are not available in pywcs-1.9: UnitConverter, and find_all_wcs(). Also, the test code seem to be not use Pythons "unittest" package? Comments and bug reports on the builds are very welcome. Cheers Ole _______________________________________________ AstroPy mailing list AstroPy at scipy.org http://mail.scipy.org/mailman/listinfo/astropy From astropy at liska.ath.cx Thu Mar 17 10:21:17 2011 From: astropy at liska.ath.cx (Ole Streicher) Date: Thu, 17 Mar 2011 15:21:17 +0100 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu> References: <4D7E39CE.907@liska.ath.cx> <4D82115A.9070409@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu> Message-ID: <4D8218DD.3060401@liska.ath.cx> Am 17.03.2011 15:00, schrieb Michael Droettboom: > The wcsset_ issue is to support compilation on MS-Windows. As wcslib > did not accept the patch upstream, we have had to maintain it in our > local copy. But that could probably be improved using #defines or > some such to make this easier for packagers to use a system copy of > wcslib. Is it OK if I patch this during the package build process? > BTW -- be careful interchanging the terms "wcslib" and "libwcs" -- > they are different projects with different licenses. Hmm, this has nothing to do with python, but can you shortly explain the differences? When I started to search for the WCS library, I found "libwcs" as part of wcstools (http://tdc-www.cfa.harvard.edu/wcstools/), and they refer to Mark Calabretta's wcslib -- and this is the one I use. Also here, to follow a bit the naming conventions (library packages should start with "lib"), I choose "libwcs" instead of "wcslib". Are there other packages to consider? Regards Ole From mdroe at stsci.edu Thu Mar 17 10:43:14 2011 From: mdroe at stsci.edu (Michael Droettboom) Date: Thu, 17 Mar 2011 14:43:14 +0000 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <4D8218DD.3060401@liska.ath.cx> References: <4D7E39CE.907@liska.ath.cx> <4D82115A.9070409@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu>, <4D8218DD.3060401@liska.ath.cx> Message-ID: <50B8371B1D275D42AE85E612144FFEB9C06196@EXCHMAIL2.stsci.edu> In SVN, I've reverted the uses of wcsset_ to wcsset, so it should work with a system wcslib without patching. On MS-Windows, some #define magic now happens to get around wcsset being a system function on that platform -- but that code path should not impact Linux. You can continue to patch it during the build process, but I think it's easier for you for us to fix this upstream. Also in SVN, I've documented how to run the tests. As you point out will not work with the latest 1.9 release, but assuming SVN head works for you for your packaging purposes, I'm happy to make a new 1.10 release. As for the library naming confusion: libwcs is from SAO/NRAO and is shipped as part of wcstools [1]. wcslib is a standalone library by Mark Calabretta [2]. The latter is what pywcs uses. But perhaps confusing is that wcstools will compile against either. pywcs only supports wcslib. I believe each of these has a common ancestor, but wcslib has been updated more recently and has more features. You may end up having to call the package "libwcslib" to avoid confusion, though perhaps "libwcs" will only ever be installed as part of "wcstools". For example, on Fedora, packages exist for Calabretta's called "wcslib", and for "wcstools", which includes libwcs. I'm not familiar with the Debian/Ubuntu world on this. Mike ________________________________________ From: astropy-bounces at scipy.org [astropy-bounces at scipy.org] on behalf of Ole Streicher [astropy at liska.ath.cx] Sent: Thursday, March 17, 2011 10:21 AM To: astropy at scipy.org Subject: Re: [AstroPy] Ubuntu packages and state of AstroLib Am 17.03.2011 15:00, schrieb Michael Droettboom: > The wcsset_ issue is to support compilation on MS-Windows. As wcslib > did not accept the patch upstream, we have had to maintain it in our > local copy. But that could probably be improved using #defines or > some such to make this easier for packagers to use a system copy of > wcslib. Is it OK if I patch this during the package build process? > BTW -- be careful interchanging the terms "wcslib" and "libwcs" -- > they are different projects with different licenses. Hmm, this has nothing to do with python, but can you shortly explain the differences? When I started to search for the WCS library, I found "libwcs" as part of wcstools (http://tdc-www.cfa.harvard.edu/wcstools/), and they refer to Mark Calabretta's wcslib -- and this is the one I use. Also here, to follow a bit the naming conventions (library packages should start with "lib"), I choose "libwcs" instead of "wcslib". Are there other packages to consider? Regards Ole _______________________________________________ AstroPy mailing list AstroPy at scipy.org http://mail.scipy.org/mailman/listinfo/astropy From sergiopr at fis.ucm.es Thu Mar 17 11:28:37 2011 From: sergiopr at fis.ucm.es (Sergio Pascual) Date: Thu, 17 Mar 2011 16:28:37 +0100 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <50B8371B1D275D42AE85E612144FFEB9C06196@EXCHMAIL2.stsci.edu> References: <4D7E39CE.907@liska.ath.cx> <4D82115A.9070409@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu> <4D8218DD.3060401@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06196@EXCHMAIL2.stsci.edu> Message-ID: 2011/3/17 Michael Droettboom : > As for the library naming confusion: ?libwcs is from SAO/NRAO and is shipped as part of wcstools [1]. ?wcslib is a standalone library by Mark Calabretta [2]. ?The latter is what pywcs uses. ?But perhaps confusing is that wcstools will compile against either. ?pywcs only supports wcslib. ?I believe each of these has a common ancestor, but wcslib has been updated more recently and has more features. ?You may end up having to call the package "libwcslib" to avoid confusion, though perhaps "libwcs" will only ever be installed as part of "wcstools". ?For example, on Fedora, packages exist for Calabretta's called "wcslib", and for "wcstools", which includes libwcs. ?I'm not familiar with the Debian/Ubuntu world on this. I would add here that wcstools is, as it name says, a set of tools (programs). The base library of these programs is called libwcs. What is very unfortunate because it's the name of the library shipped with wcslib (there's an open bug in Fedora about that). wcstools contains code of some old version of wcslib. Sergio -- Sergio Pascual ? ? ? ? ? ? ? ? ? ? ? ?http://guaix.fis.ucm.es/~spr gpg fingerprint: 5203 B42D 86A0 5649 410A F4AC A35F D465 F263 BCCC Departamento de Astrof?sica -- Universidad Complutense de Madrid (Spain) From astropy at liska.ath.cx Thu Mar 17 11:37:01 2011 From: astropy at liska.ath.cx (Ole Streicher) Date: Thu, 17 Mar 2011 16:37:01 +0100 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <50B8371B1D275D42AE85E612144FFEB9C06196@EXCHMAIL2.stsci.edu> References: <4D7E39CE.907@liska.ath.cx> <4D82115A.9070409@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu> <4D8218DD.3060401@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06196@EXCHMAIL2.stsci.edu> Message-ID: <4D822A9D.6050803@liska.ath.cx> Am 17.03.2011 15:43, schrieb Michael Droettboom: > Also in SVN, I've documented how to run the tests. As you point out > will not work with the latest 1.9 release, but assuming SVN head > works for you for your packaging purposes, I'm happy to make a new > 1.10 release. I successfully tested the package build process against the SVN trunk, so a new version would be fine for me. Additionally to Sergio's patch, I will also remove all references to the test fits files and the wcslib includes in the defsetup.py, since they are not needed for the binary package. This makes the package smaller without the need to remove the tests from the source tarball. Would you just tag the version in SVN or also provide a downloadable tar file at stsci? Best Ole -------------- next part -------------- A non-text attachment was scrubbed... Name: pywcs-unbundle.patch Type: text/x-diff Size: 1316 bytes Desc: not available URL: From astropy at liska.ath.cx Thu Mar 17 11:55:57 2011 From: astropy at liska.ath.cx (Ole Streicher) Date: Thu, 17 Mar 2011 16:55:57 +0100 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: References: <4D7E39CE.907@liska.ath.cx> <4D82115A.9070409@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu> <4D8218DD.3060401@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06196@EXCHMAIL2.stsci.edu> Message-ID: <4D822F0D.5070102@liska.ath.cx> Am 17.03.2011 16:28, schrieb Sergio Pascual: > 2011/3/17 Michael Droettboom : >> As for the library naming confusion: libwcs is from SAO/NRAO and >> is shipped as part of wcstools [1]. wcslib is a standalone library >> by Mark Calabretta [2]. The latter is what pywcs uses. But >> perhaps confusing is that wcstools will compile against either. >> pywcs only supports wcslib. I believe each of these has a common >> ancestor, but wcslib has been updated more recently and has more >> features. You may end up having to call the package "libwcslib" to >> avoid confusion, though perhaps "libwcs" will only ever be >> installed as part of "wcstools". For example, on Fedora, packages >> exist for Calabretta's called "wcslib", and for "wcstools", which >> includes libwcs. I'm not familiar with the Debian/Ubuntu world on >> this. > > I would add here that wcstools is, as it name says, a set of tools > (programs). The base library of these programs is called libwcs. > What is very unfortunate because it's the name of the library shipped > with wcslib (there's an open bug in Fedora about that). wcstools > contains code of some old version of wcslib. No I am /really/ lost. - wcslib-4.7 (or 4.4.4) is a separate packge maintained by Mark Calabretta, installs a lib "libwcs.a" (and .so's) in /usr/lib/ - wcstools-3.8.1 is a complete WCS utility toolkit that comes with its own "libwcs", that also compiles as static "libwcs.a". However, this library is not updated since 2 years. - wcstools-3.8.1 may use Mark's wcslib-4.X instead of the builtin version (statement on the wcstools homepage) >From there, I would not see any need to keep the wcstools-3.8.1/libwcs since it is superceded by wcslib, except maybe for some legacy code. If "wcstools-3.8.1/libwcs" is just an older version of "wcslib-4.7", one would probably better opt for adopting wcslib-4.7 instead of using the old library in the packaging process. If this does not work, one can still provide the old library as "liblegacywcs" or similar. Is there anything missing in this scenario? Cheers Ole From sergiopr at fis.ucm.es Thu Mar 17 13:32:18 2011 From: sergiopr at fis.ucm.es (Sergio Pascual) Date: Thu, 17 Mar 2011 18:32:18 +0100 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <4D822F0D.5070102@liska.ath.cx> References: <4D7E39CE.907@liska.ath.cx> <4D82115A.9070409@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu> <4D8218DD.3060401@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06196@EXCHMAIL2.stsci.edu> <4D822F0D.5070102@liska.ath.cx> Message-ID: 2011/3/17 Ole Streicher : > Am 17.03.2011 16:28, schrieb Sergio Pascual: >> 2011/3/17 Michael Droettboom : >>> As for the library naming confusion: ?libwcs is from SAO/NRAO and >>> is shipped as part of wcstools [1]. ?wcslib is a standalone library >>> by Mark Calabretta [2]. ?The latter is what pywcs uses. ?But >>> perhaps confusing is that wcstools will compile against either. >>> pywcs only supports wcslib. ?I believe each of these has a common >>> ancestor, but wcslib has been updated more recently and has more >>> features. ?You may end up having to call the package "libwcslib" to >>> avoid confusion, though perhaps "libwcs" will only ever be >>> installed as part of "wcstools". ?For example, on Fedora, packages >>> exist for Calabretta's called "wcslib", and for "wcstools", which >>> includes libwcs. ?I'm not familiar with the Debian/Ubuntu world on >>> this. >> >> I would add here that wcstools is, as it name says, a set of tools >> (programs). The base library of these programs is called libwcs. >> What is very unfortunate because it's the name of the library shipped >> with wcslib (there's an open bug in Fedora about that). wcstools >> contains code of some old version of wcslib. > > No I am /really/ lost. > > - wcslib-4.7 (or 4.4.4) is a separate packge maintained by Mark > ?Calabretta, installs a lib "libwcs.a" (and .so's) in /usr/lib/ > > - wcstools-3.8.1 is a complete WCS utility toolkit that comes with > ?its own "libwcs", that also compiles as static "libwcs.a". However, > ?this library is not updated since 2 years. > > - wcstools-3.8.1 may use Mark's wcslib-4.X instead of the builtin > ?version (statement on the wcstools homepage) > > >From there, I would not see any need to keep the wcstools-3.8.1/libwcs > since it is superceded by wcslib, except maybe for some legacy code. If > "wcstools-3.8.1/libwcs" is just an older version of "wcslib-4.7", one > would probably better opt for adopting wcslib-4.7 instead of using the > old library in the packaging process. If this does not work, one can > still provide the old library as "liblegacywcs" or similar. Is there > anything missing in this scenario? To put it in a few words: some parts of wcslib are in wcstools. There are a lot more things that are in wcstools and not in wcslib (functions related with astrometry, catalogue search, etc.) I included wcstools in Fedora only because is used by ds9, but the name colision with wcslib is very inconvenient. In your case, I would forget about wcstools Regards > > Cheers > > Ole > > _______________________________________________ > AstroPy mailing list > AstroPy at scipy.org > http://mail.scipy.org/mailman/listinfo/astropy > -- Sergio Pascual ? ? ? ? ? ? ? ? ? ? ? ?http://guaix.fis.ucm.es/~spr gpg fingerprint: 5203 B42D 86A0 5649 410A F4AC A35F D465 F263 BCCC Departamento de Astrof?sica -- Universidad Complutense de Madrid (Spain) From mdroe at stsci.edu Thu Mar 17 15:01:59 2011 From: mdroe at stsci.edu (Michael Droettboom) Date: Thu, 17 Mar 2011 19:01:59 +0000 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <4D822A9D.6050803@liska.ath.cx> References: <4D7E39CE.907@liska.ath.cx> <4D82115A.9070409@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu> <4D8218DD.3060401@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06196@EXCHMAIL2.stsci.edu>, <4D822A9D.6050803@liska.ath.cx> Message-ID: <50B8371B1D275D42AE85E612144FFEB9C06361@EXCHMAIL2.stsci.edu> I have made a pywcs-1.10-4.7 release, available here: http://stsdas.stsci.edu/astrolib/pywcs-1.10-4.7.tar.gz and tagged in SVN here: http://svn6.assembla.com/svn/astrolib/tags/pywcs_1.10/ That should remove the need for the crazy wcsset_ patching, and also provide some basic unit tests for running with nose. Cheers, Mike ________________________________________ From: astropy-bounces at scipy.org [astropy-bounces at scipy.org] on behalf of Ole Streicher [astropy at liska.ath.cx] Sent: Thursday, March 17, 2011 11:37 AM Cc: astropy at scipy.org Subject: Re: [AstroPy] Ubuntu packages and state of AstroLib Am 17.03.2011 15:43, schrieb Michael Droettboom: > Also in SVN, I've documented how to run the tests. As you point out > will not work with the latest 1.9 release, but assuming SVN head > works for you for your packaging purposes, I'm happy to make a new > 1.10 release. I successfully tested the package build process against the SVN trunk, so a new version would be fine for me. Additionally to Sergio's patch, I will also remove all references to the test fits files and the wcslib includes in the defsetup.py, since they are not needed for the binary package. This makes the package smaller without the need to remove the tests from the source tarball. Would you just tag the version in SVN or also provide a downloadable tar file at stsci? Best Ole From erik.tollerud at gmail.com Thu Mar 17 18:16:09 2011 From: erik.tollerud at gmail.com (Erik Tollerud) Date: Thu, 17 Mar 2011 15:16:09 -0700 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <50B8371B1D275D42AE85E612144FFEB9C06361@EXCHMAIL2.stsci.edu> References: <4D7E39CE.907@liska.ath.cx> <4D82115A.9070409@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu> <4D8218DD.3060401@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06196@EXCHMAIL2.stsci.edu> <4D822A9D.6050803@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06361@EXCHMAIL2.stsci.edu> Message-ID: So are there still no plans to upload pywcs to PyPI? I can understand not wanting to deal with building binaries, but if a source distribution can be uploaded, then the "pip" tool (which is the now-standard tool given that easy_install is not being developed nor is it py3k compatible) can properly install from source as long as the user has wcslib installed. This makes it much easier to treat pywcs as a dependency for other packages. As far as I can tell this should be as easy as adding " 'requires':['numpy','pyfits'],'provides':['pywcs']" to the setupargs in defsetup.py and running "python sdist upload". Presumably numpy is still needed to build, but as long as numpy is installed, I think that should make it so that other packages can make pywcs a dependency. On Thu, Mar 17, 2011 at 12:01 PM, Michael Droettboom wrote: > I have made a pywcs-1.10-4.7 release, available here: > > http://stsdas.stsci.edu/astrolib/pywcs-1.10-4.7.tar.gz > > and tagged in SVN here: > > http://svn6.assembla.com/svn/astrolib/tags/pywcs_1.10/ > > That should remove the need for the crazy wcsset_ patching, and also provide some basic unit tests for running with nose. > > Cheers, > Mike > > ________________________________________ > From: astropy-bounces at scipy.org [astropy-bounces at scipy.org] on behalf of Ole Streicher [astropy at liska.ath.cx] > Sent: Thursday, March 17, 2011 11:37 AM > Cc: astropy at scipy.org > Subject: Re: [AstroPy] Ubuntu packages and state of AstroLib > > Am 17.03.2011 15:43, schrieb Michael Droettboom: >> Also in SVN, I've documented how to run the tests. ?As you point out >> will not work with the latest 1.9 release, but assuming SVN head >> works for you for your packaging purposes, I'm happy to make a new >> 1.10 release. > > I successfully tested the package build process against the SVN trunk, > so a new version would be fine for me. > > Additionally to Sergio's patch, I will also remove all references to the > test fits files and the wcslib includes in the defsetup.py, since they > are not needed for the binary package. This makes the package smaller > without the need to remove the tests from the source tarball. > > Would you just tag the version in SVN or also provide a downloadable tar > file at stsci? > > Best > > Ole > _______________________________________________ > AstroPy mailing list > AstroPy at scipy.org > http://mail.scipy.org/mailman/listinfo/astropy > -- Erik Tollerud From mdroe at stsci.edu Thu Mar 17 19:02:55 2011 From: mdroe at stsci.edu (Michael Droettboom) Date: Thu, 17 Mar 2011 23:02:55 +0000 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: References: <4D7E39CE.907@liska.ath.cx> <4D82115A.9070409@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu> <4D8218DD.3060401@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06196@EXCHMAIL2.stsci.edu> <4D822A9D.6050803@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06361@EXCHMAIL2.stsci.edu>, Message-ID: <50B8371B1D275D42AE85E612144FFEB9C06666@EXCHMAIL2.stsci.edu> I've uploaded a 1.10 source release to PyPI. Please let me know how it works, as I've never really used easy_install/pip. Mike ________________________________________ From: etollerud at gmail.com [etollerud at gmail.com] on behalf of Erik Tollerud [erik.tollerud at gmail.com] Sent: Thursday, March 17, 2011 6:16 PM To: Michael Droettboom; astropy at scipy.org Subject: Re: [AstroPy] Ubuntu packages and state of AstroLib So are there still no plans to upload pywcs to PyPI? I can understand not wanting to deal with building binaries, but if a source distribution can be uploaded, then the "pip" tool (which is the now-standard tool given that easy_install is not being developed nor is it py3k compatible) can properly install from source as long as the user has wcslib installed. This makes it much easier to treat pywcs as a dependency for other packages. As far as I can tell this should be as easy as adding " 'requires':['numpy','pyfits'],'provides':['pywcs']" to the setupargs in defsetup.py and running "python sdist upload". Presumably numpy is still needed to build, but as long as numpy is installed, I think that should make it so that other packages can make pywcs a dependency. On Thu, Mar 17, 2011 at 12:01 PM, Michael Droettboom wrote: > I have made a pywcs-1.10-4.7 release, available here: > > http://stsdas.stsci.edu/astrolib/pywcs-1.10-4.7.tar.gz > > and tagged in SVN here: > > http://svn6.assembla.com/svn/astrolib/tags/pywcs_1.10/ > > That should remove the need for the crazy wcsset_ patching, and also provide some basic unit tests for running with nose. > > Cheers, > Mike > > ________________________________________ > From: astropy-bounces at scipy.org [astropy-bounces at scipy.org] on behalf of Ole Streicher [astropy at liska.ath.cx] > Sent: Thursday, March 17, 2011 11:37 AM > Cc: astropy at scipy.org > Subject: Re: [AstroPy] Ubuntu packages and state of AstroLib > > Am 17.03.2011 15:43, schrieb Michael Droettboom: >> Also in SVN, I've documented how to run the tests. As you point out >> will not work with the latest 1.9 release, but assuming SVN head >> works for you for your packaging purposes, I'm happy to make a new >> 1.10 release. > > I successfully tested the package build process against the SVN trunk, > so a new version would be fine for me. > > Additionally to Sergio's patch, I will also remove all references to the > test fits files and the wcslib includes in the defsetup.py, since they > are not needed for the binary package. This makes the package smaller > without the need to remove the tests from the source tarball. > > Would you just tag the version in SVN or also provide a downloadable tar > file at stsci? > > Best > > Ole > _______________________________________________ > AstroPy mailing list > AstroPy at scipy.org > http://mail.scipy.org/mailman/listinfo/astropy > -- Erik Tollerud From erik.tollerud at gmail.com Thu Mar 17 19:21:52 2011 From: erik.tollerud at gmail.com (Erik Tollerud) Date: Thu, 17 Mar 2011 16:21:52 -0700 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <50B8371B1D275D42AE85E612144FFEB9C06666@EXCHMAIL2.stsci.edu> References: <4D7E39CE.907@liska.ath.cx> <4D82115A.9070409@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu> <4D8218DD.3060401@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06196@EXCHMAIL2.stsci.edu> <4D822A9D.6050803@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06361@EXCHMAIL2.stsci.edu> <50B8371B1D275D42AE85E612144FFEB9C06666@EXCHMAIL2.stsci.edu> Message-ID: I just tested in in an ubuntu virtualenv, and it seems to work. I was forced to also install pyfits, apparently because the pywcs setup script uses just distutils instead of distutils+distribute (or distutils+setuptools). That's not a serious problem, though, as it's trivial to install with PIP if someone needs it. Thanks for posting it! On Thu, Mar 17, 2011 at 4:02 PM, Michael Droettboom wrote: > I've uploaded a 1.10 source release to PyPI. ?Please let me know how it works, as I've never really used easy_install/pip. > > Mike > > ________________________________________ > From: etollerud at gmail.com [etollerud at gmail.com] on behalf of Erik Tollerud [erik.tollerud at gmail.com] > Sent: Thursday, March 17, 2011 6:16 PM > To: Michael Droettboom; astropy at scipy.org > Subject: Re: [AstroPy] Ubuntu packages and state of AstroLib > > So are there still no plans to upload pywcs to PyPI? > > I can understand not wanting to deal with building binaries, but if a > source distribution can be uploaded, then the "pip" tool (which is the > now-standard tool given that easy_install is not being developed nor > is it py3k compatible) can properly install from source as long as the > user has wcslib installed. ?This makes it much easier to treat pywcs > as a dependency for other packages. > > As far as I can tell this should be as easy as adding " > 'requires':['numpy','pyfits'],'provides':['pywcs']" to the setupargs > in defsetup.py and running "python sdist upload". Presumably numpy is > still needed to build, but as long as numpy is installed, I think that > should make it so that other packages can make pywcs a dependency. > > > On Thu, Mar 17, 2011 at 12:01 PM, Michael Droettboom wrote: >> I have made a pywcs-1.10-4.7 release, available here: >> >> http://stsdas.stsci.edu/astrolib/pywcs-1.10-4.7.tar.gz >> >> and tagged in SVN here: >> >> http://svn6.assembla.com/svn/astrolib/tags/pywcs_1.10/ >> >> That should remove the need for the crazy wcsset_ patching, and also provide some basic unit tests for running with nose. >> >> Cheers, >> Mike >> >> ________________________________________ >> From: astropy-bounces at scipy.org [astropy-bounces at scipy.org] on behalf of Ole Streicher [astropy at liska.ath.cx] >> Sent: Thursday, March 17, 2011 11:37 AM >> Cc: astropy at scipy.org >> Subject: Re: [AstroPy] Ubuntu packages and state of AstroLib >> >> Am 17.03.2011 15:43, schrieb Michael Droettboom: >>> Also in SVN, I've documented how to run the tests. ?As you point out >>> will not work with the latest 1.9 release, but assuming SVN head >>> works for you for your packaging purposes, I'm happy to make a new >>> 1.10 release. >> >> I successfully tested the package build process against the SVN trunk, >> so a new version would be fine for me. >> >> Additionally to Sergio's patch, I will also remove all references to the >> test fits files and the wcslib includes in the defsetup.py, since they >> are not needed for the binary package. This makes the package smaller >> without the need to remove the tests from the source tarball. >> >> Would you just tag the version in SVN or also provide a downloadable tar >> file at stsci? >> >> Best >> >> Ole >> _______________________________________________ >> AstroPy mailing list >> AstroPy at scipy.org >> http://mail.scipy.org/mailman/listinfo/astropy >> > > > > -- > Erik Tollerud > -- Erik Tollerud From trice at cfa.harvard.edu Fri Mar 18 01:19:27 2011 From: trice at cfa.harvard.edu (Tom Rice) Date: Fri, 18 Mar 2011 01:19:27 -0400 Subject: [AstroPy] Robust Statistics in Numpy? Message-ID: Hi all, I'm looking for a package in Python / Numpy / SciPy that provides robust statistical functions, such as the Resistant Mean or the Robust standard deviation - essentially, a Python analogue of the IDL Robust Statistics package described here: http://idlastro.gsfc.nasa.gov/contents.html#C17 http://idlastro.gsfc.nasa.gov/ftp/pro/robust/aaareadme.txt After some googling, I found an ad hoc implementation of two of these (robust mean and robust standard deviation) at this website: http://code.google.com/p/fmccd/source/browse/trunk/fermiMCCD.py?spec=svn18&r=18 but I'm interested in an actual, maintained package, if one exists (and I feel that there should be one). Thanks very much, Tom Rice -- Thomas S. Rice trice at cfa.harvard.edu -------------- next part -------------- An HTML attachment was scrubbed... URL: From astropy at liska.ath.cx Fri Mar 18 05:04:24 2011 From: astropy at liska.ath.cx (Ole Streicher) Date: Fri, 18 Mar 2011 10:04:24 +0100 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: <50B8371B1D275D42AE85E612144FFEB9C06361@EXCHMAIL2.stsci.edu> References: <4D7E39CE.907@liska.ath.cx> <4D82115A.9070409@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu> <4D8218DD.3060401@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06196@EXCHMAIL2.stsci.edu> <4D822A9D.6050803@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06361@EXCHMAIL2.stsci.edu> Message-ID: <4D832018.9060200@liska.ath.cx> I put python-pywcs-1.10 to my Ubuntu repository. Thank you for the new version! Cheers Ole Am 17.03.2011 20:01, schrieb Michael Droettboom: > I have made a pywcs-1.10-4.7 release, available here: > > http://stsdas.stsci.edu/astrolib/pywcs-1.10-4.7.tar.gz > > and tagged in SVN here: > > http://svn6.assembla.com/svn/astrolib/tags/pywcs_1.10/ > > That should remove the need for the crazy wcsset_ patching, and also > provide some basic unit tests for running with nose. > > Cheers, Mike From johann.cohentanugi at gmail.com Fri Mar 18 08:38:23 2011 From: johann.cohentanugi at gmail.com (Johann Cohen-Tanugi) Date: Fri, 18 Mar 2011 13:38:23 +0100 Subject: [AstroPy] Robust Statistics in Numpy? In-Reply-To: References: Message-ID: <4D83523F.3090406@gmail.com> hi Tom, this is probably a request best answered if channeled to the scipy users mailing list. Johann On 03/18/2011 06:19 AM, Tom Rice wrote: > Hi all, > > I'm looking for a package in Python / Numpy / SciPy that provides > robust statistical functions, such as the Resistant Mean or the Robust > standard deviation - essentially, a Python analogue of the IDL Robust > Statistics package described here: > http://idlastro.gsfc.nasa.gov/contents.html#C17 > http://idlastro.gsfc.nasa.gov/ftp/pro/robust/aaareadme.txt > > After some googling, I found an ad hoc implementation of two of these > (robust mean and robust standard deviation) at this website: > http://code.google.com/p/fmccd/source/browse/trunk/fermiMCCD.py?spec=svn18&r=18 > > but I'm interested in an actual, maintained package, if one exists > (and I feel that there should be one). > > Thanks very much, > Tom Rice > > -- > Thomas S. Rice > trice at cfa.harvard.edu > > > -- > This message has been scanned for viruses and > dangerous content by *MailScanner* , and is > believed to be clean. > > > _______________________________________________ > AstroPy mailing list > AstroPy at scipy.org > http://mail.scipy.org/mailman/listinfo/astropy -------------- next part -------------- An HTML attachment was scrubbed... URL: From embray at stsci.edu Fri Mar 18 10:05:48 2011 From: embray at stsci.edu (Erik Bray) Date: Fri, 18 Mar 2011 10:05:48 -0400 Subject: [AstroPy] Ubuntu packages and state of AstroLib In-Reply-To: References: <4D7E39CE.907@liska.ath.cx> <4D82115A.9070409@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C0616B@EXCHMAIL2.stsci.edu> <4D8218DD.3060401@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06196@EXCHMAIL2.stsci.edu> <4D822A9D.6050803@liska.ath.cx> <50B8371B1D275D42AE85E612144FFEB9C06361@EXCHMAIL2.stsci.edu> Message-ID: <4D8366BC.6070404@stsci.edu> Erik Tollerud wrote: > So are there still no plans to upload pywcs to PyPI? > As far as I can tell this should be as easy as adding " > 'requires':['numpy','pyfits'],'provides':['pywcs']" to the setupargs > in defsetup.py and running "python sdist upload". More or less that's true. Though the distutils hacks that pyfits, pywcs, and some of our other packages rely on makes doing this "right" a bit trickier. I'm in the process of overhauling these packages so that it will be easier to distribute them on PyPI (with the correct dependency management) and have releases pushed out regularly. Erik From sienkiew at stsci.edu Fri Mar 18 12:04:29 2011 From: sienkiew at stsci.edu (Mark Sienkiewicz) Date: Fri, 18 Mar 2011 12:04:29 -0400 Subject: [AstroPy] Astrolib release coming Message-ID: <4D83828D.9060505@stsci.edu> Obviously you can tell that we have not given much attention to making an Astrolib release, but we are planning to change that. The release will come in the form of an upload to pypi (so you can install it with pip or easyinstall) and a source tarball. It will take some modifications to do this, but we expect to make the release some time in the next few weeks. If you are an author of a package in Astrolib, we need to know if you want to do anything to your code before the release. If you do, contact me or discuss it on the list. Mark S From amenity at enthought.com Wed Mar 23 08:56:17 2011 From: amenity at enthought.com (Amenity Applewhite) Date: Wed, 23 Mar 2011 07:56:17 -0500 Subject: [AstroPy] SciPy 2011 Call for Papers Message-ID: Hello, SciPy 2011 , the 10th Python in Science conference, will be held July 11 - 16, 2011, in Austin, TX. At this conference, novel applications and breakthroughs made in the pursuit of science using Python are presented. Attended by leading figures from both academia and industry, it is an excellent opportunity to experience the cutting edge of scientific software development. The conference is preceded by two days of tutorials, during which community experts provide training on several scientific Python packages. *We'd like to invite you to consider presenting at SciPy 2011.* The list of topics that are appropriate for the conference includes (but is not limited to): * new Python libraries for science and engineering; * applications of Python to the solution of scientific or computational problems; * high performance, parallel and GPU computing with Python; * use of Python in science education. *Specialized Tracks* This year we also have two specialized tracks. They will be run concurrent to the main conference. *Python in Data Science Chair: Peter Wang, Streamitive, Inc.* This track focuses on the advantages and challenges of applying Python in the emerging field of "data science". This includes a breadth of technologies, from wrangling realtime data streams from the social web, to machine learning and semantic analysis, to workflow and repository management for large datasets. *Python and Core Technologies Chair: Anthony Scopatz, Enthought, Inc.* In an effort to broaden the scope of SciPy and to engage the larger community of software developers, we are pleased to introduce the _Python & Core Technologies_ track. Talks will cover subjects that are not directly related to science and engineering, yet nonetheless affect scientific computing. Proposals on the Python language, visualization toolkits, web frameworks, education, and other topics are appropriate for this session. *Talk/Paper Submission* We invite you to take part by submitting a talk abstract on the conference website at: http://conference.scipy.org/scipy2011/papers.php Papers are included in the peer-reviewed conference proceedings, to be published online. *Important dates for authors:* Friday, April 15: Tutorial proposals due (remember: stipends will be provided for Tutorial instructors) http://conference.scipy.org/scipy2011/tutorials.php Sunday, April 24: Paper abstracts due Sunday, May 8: Student sponsorship request due http://conference.scipy.org/scipy2011/student.php Tuesday, May 10: Accepted talks announced Monday, May 16: Student sponsorships announced Monday, May 23: Early Registration ends Sunday, June 20: Papers due Monday-Tuesday, July 11 - 12: Tutorials Wednesday-Thursday, July 13 - July 14: Conference Friday-Saturday, July 15 - July 16: Sprints The SciPy 2011 Team @SciPy2011 http://twitter.com/SciPy2011 _________________________ Amenity Applewhite Enthought, Inc. Scientific Computing Solutions -------------- next part -------------- An HTML attachment was scrubbed... URL: From yogesh at ncra.tifr.res.in Wed Mar 23 09:23:34 2011 From: yogesh at ncra.tifr.res.in (Yogesh Wadadekar) Date: Wed, 23 Mar 2011 18:53:34 +0530 (IST) Subject: [AstroPy] pywcs issue Message-ID: I am trying to use pywcs for analysis of some FIRST survey radio data. When I run the simple python test script (pasted at the bottom of this email) below on the FITS file: 00525+00130Z.fits Note that the FITS file above has minimal WCS related headers extracted from the original FIRST data. It is available at: http://www.ncra.tifr.res.in/~yogesh/00525+00130Z.fits I get strange results. The output of the script is: project at yogesh:~$ python test.py running pywcs version:1.10-4.7 Image: 00525+00130Z.fits Input ra and dec: 12.742937 0.263592 Ra and dec after conversion via wcs_sky2pix and wcs_pix2sky: [ 12.742937] [ 0.263592] Input ra and dec: 192.983075 -0.1740946 Ra and dec after conversion via wcs_sky2pix and wcs_pix2sky: [ 13.26637025] [ 0.25923741] For a RA/Dec lying within the image, the sky --> pix ---> sky conversion gives correct results, but for a position ~180 degrees away, wcs_sky2pix returns x,y within the image. wcs_pix2sky happily converts this x,y to a grossly incorrect RA,Dec. I am completely perplexed about what is going on here. Yogesh Yogesh Wadadekar Tel: 91 20 25719238 National Centre for Radio Astrophysics e-mail: yogesh at ncra.tifr.res.in Post Bag 3, Ganeshkhind Pune 411007, India ------------------------------------------------------------------------ #!/usr/bin/env python import pywcs,pyfits print 'running pywcs version:'+pywcs.__version__ image = '00525+00130Z.fits' print 'Image: ',image header = pyfits.getheader(image) wcs= pywcs.WCS(header) x, y = wcs.wcs_sky2pix(12.742937,0.263592,1) ra, dec = wcs.wcs_pix2sky(x, y,1) print 'Input ra and dec:\n',12.742937,'\t',0.263592,'\nRa and dec after conversion via wcs_sky2pix and wcs_pix2sky:\n',ra,'\t',dec # Now repeat with ra,dec pair for a position well outside the image x, y = wcs.wcs_sky2pix(192.983075,-0.1740946,1) ra, dec = wcs.wcs_pix2sky(x, y,1) print 'Input ra and dec:\n',192.983075,'\t', -0.1740946,'\nRa and dec after conversion via wcs_sky2pix and wcs_pix2sky:\n',ra,'\t',dec From mdroe at stsci.edu Thu Mar 24 12:18:43 2011 From: mdroe at stsci.edu (Michael Droettboom) Date: Thu, 24 Mar 2011 16:18:43 +0000 Subject: [AstroPy] pywcs issue In-Reply-To: References: Message-ID: <50B8371B1D275D42AE85E612144FFEB9C1AB15@EXCHMAIL1.stsci.edu> The WCS is not guaranteed to be defined outside of the image. Particularly with SIN projections, it is often the case that the results are invalid when outside of the image. wcslib (and pywcs) does not do any sanity checks for this. Mike ________________________________________ From: astropy-bounces at scipy.org [astropy-bounces at scipy.org] on behalf of Yogesh Wadadekar [yogesh at ncra.tifr.res.in] Sent: Wednesday, March 23, 2011 9:23 AM To: astropy at scipy.org Cc: ishan.07 at iist.ac.in Subject: [AstroPy] pywcs issue I am trying to use pywcs for analysis of some FIRST survey radio data. When I run the simple python test script (pasted at the bottom of this email) below on the FITS file: 00525+00130Z.fits Note that the FITS file above has minimal WCS related headers extracted from the original FIRST data. It is available at: http://www.ncra.tifr.res.in/~yogesh/00525+00130Z.fits I get strange results. The output of the script is: project at yogesh:~$ python test.py running pywcs version:1.10-4.7 Image: 00525+00130Z.fits Input ra and dec: 12.742937 0.263592 Ra and dec after conversion via wcs_sky2pix and wcs_pix2sky: [ 12.742937] [ 0.263592] Input ra and dec: 192.983075 -0.1740946 Ra and dec after conversion via wcs_sky2pix and wcs_pix2sky: [ 13.26637025] [ 0.25923741] For a RA/Dec lying within the image, the sky --> pix ---> sky conversion gives correct results, but for a position ~180 degrees away, wcs_sky2pix returns x,y within the image. wcs_pix2sky happily converts this x,y to a grossly incorrect RA,Dec. I am completely perplexed about what is going on here. Yogesh Yogesh Wadadekar Tel: 91 20 25719238 National Centre for Radio Astrophysics e-mail: yogesh at ncra.tifr.res.in Post Bag 3, Ganeshkhind Pune 411007, India ------------------------------------------------------------------------ #!/usr/bin/env python import pywcs,pyfits print 'running pywcs version:'+pywcs.__version__ image = '00525+00130Z.fits' print 'Image: ',image header = pyfits.getheader(image) wcs= pywcs.WCS(header) x, y = wcs.wcs_sky2pix(12.742937,0.263592,1) ra, dec = wcs.wcs_pix2sky(x, y,1) print 'Input ra and dec:\n',12.742937,'\t',0.263592,'\nRa and dec after conversion via wcs_sky2pix and wcs_pix2sky:\n',ra,'\t',dec # Now repeat with ra,dec pair for a position well outside the image x, y = wcs.wcs_sky2pix(192.983075,-0.1740946,1) ra, dec = wcs.wcs_pix2sky(x, y,1) print 'Input ra and dec:\n',192.983075,'\t', -0.1740946,'\nRa and dec after conversion via wcs_sky2pix and wcs_pix2sky:\n',ra,'\t',dec _______________________________________________ AstroPy mailing list AstroPy at scipy.org http://mail.scipy.org/mailman/listinfo/astropy From yogesh at ncra.tifr.res.in Fri Mar 25 13:03:24 2011 From: yogesh at ncra.tifr.res.in (Yogesh Wadadekar) Date: Fri, 25 Mar 2011 22:33:24 +0530 (IST) Subject: [AstroPy] pywcs issue In-Reply-To: <50B8371B1D275D42AE85E612144FFEB9C1AB15@EXCHMAIL1.stsci.edu> References: <50B8371B1D275D42AE85E612144FFEB9C1AB15@EXCHMAIL1.stsci.edu> Message-ID: > The WCS is not guaranteed to be defined outside of the image. > Particularly with SIN projections, it is often the case that the results > are invalid when outside of the image. wcslib (and pywcs) does not do > any sanity checks for this. Thank you for this clarification. One expects that the conversions will become progressively more inaccurate as one moves further out from the image. But it is strange that at the greatest possible distance away on the sphere (~180 degrees away), results are similar to what one sees at very small distances from the image center. I will check the SIN projection equations to get a feel for this; I have not seen this happen the more commonly used TAN projection. We will, of course, use our own sanity checks to operationally overcome the problem. reg, Yogesh Yogesh Wadadekar Tel: 91 20 25719238 National Centre for Radio Astrophysics e-mail: yogesh at ncra.tifr.res.in Post Bag 3, Ganeshkhind Pune 411007, India From jslavin at cfa.harvard.edu Wed Mar 30 09:18:05 2011 From: jslavin at cfa.harvard.edu (Jonathan Slavin) Date: Wed, 30 Mar 2011 09:18:05 -0400 Subject: [AstroPy] support for multi-dimensional table columns in pyfits Message-ID: <1301491085.31252.15.camel@shevek> To any pyfits developers out there: I'd like to request that support for multidimensional columns in binary tables be added to pyfits. It doesn't seem that it should be too hard, though in going through the code it's not clear to me the best way to proceed -- the Delayed type of data complicates things. The dimensions that the array should have are contained in the dims attribute of the Column object (as a string), but multidimensional arrays are returned as 1-D arrays. Once reshaped, the data is correct (though the order of the dimensions has to be reversed). The IDL routine mrdfits handles this very easily. Ideally one could do something like: data = pyfits.getdata(file,exten) and data.varname would contain an ndarray of the proper shape. As it is, the returned variables have shape (1,N) where N is the total number of elements. Jon -- ______________________________________________________________ Jonathan D. Slavin Harvard-Smithsonian CfA jslavin at cfa.harvard.edu 60 Garden Street, MS 83 phone: (617) 496-7981 Cambridge, MA 02138-1516 cell: (781) 363-0035 USA ______________________________________________________________ From erin.sheldon at gmail.com Wed Mar 30 09:50:46 2011 From: erin.sheldon at gmail.com (Erin Sheldon) Date: Wed, 30 Mar 2011 07:50:46 -0600 Subject: [AstroPy] support for multi-dimensional table columns in pyfits In-Reply-To: <1301491085.31252.15.camel@shevek> References: <1301491085.31252.15.camel@shevek> Message-ID: On Wed, Mar 30, 2011 at 7:18 AM, Jonathan Slavin wrote: > To any pyfits developers out there: > > I'd like to request that support for multidimensional columns in binary > tables be added to pyfits. Jonathan- I actually recently requested this very feature, and Erik Bray of space telescope is working on it. It is almost ready. -e -- Erin Scott Sheldon Brookhaven National Laboratory From embray at stsci.edu Wed Mar 30 12:24:55 2011 From: embray at stsci.edu (Erik Bray) Date: Wed, 30 Mar 2011 12:24:55 -0400 Subject: [AstroPy] support for multi-dimensional table columns in pyfits In-Reply-To: References: <1301491085.31252.15.camel@shevek> Message-ID: <4D935957.9080709@stsci.edu> Erin Sheldon wrote: > On Wed, Mar 30, 2011 at 7:18 AM, Jonathan Slavin > wrote: >> To any pyfits developers out there: >> >> I'd like to request that support for multidimensional columns in binary >> tables be added to pyfits. > > Jonathan- > > I actually recently requested this very feature, and Erik Bray of space > telescope is working on it. It is almost ready. Right; this is mostly working in the development branch of pyfits right now, and should be in the next release. As Erin wrote, there are still a couple problems with the implementation as is, though multi-dimensional columns should be returned in the correct shape. Thanks, Erik From sbellem at gmail.com Thu Mar 31 01:27:32 2011 From: sbellem at gmail.com (Sylvain Bellemare) Date: Wed, 30 Mar 2011 22:27:32 -0700 Subject: [AstroPy] Willing to contribute to SciPy & NumPy ... Message-ID: Hi, I would like to seriously start contributing to NumPy and/or SciPy, as much as I possibly can. Because NumPy & SciPy are new to me, I'm not so sure whether I could be of some help. But if my help is welcomed, and if it makes sense to ask: "What could I best contribute to?", then perhaps some background information about myself would be useful ... *ACADEMICS* I completed an undergraduate degree in Computer Science, along with some mathematics and physics courses. For the mathematics, I did some Calculus, Linear Algebra, Differential Equations, Fourier series, Fourier & Laplace transform. In Physics, I did up to and including a first course in Quantum Mechanics. I also did some undergraduate research work, that involved computational high energy physics (proton-proton collision). The work was done with some Mathematica packages (FeynArts ). *INDUSTRY* I currently work in the software industry and for the past few months, have been performing both software development and system administration tasks. In terms of object-oriented programming, my main experience is with Java. I'm a beginner with Python. I used Python at work to automate the reverse-engineering process of some Java programs. Career-wise, one of my goals is to go into scientific computing, and use Python as the main programming language, and to perhaps pursue graduate studies in Computational Physics, hence my current interest for contributing to NumPy & SciPy. If you wish to know more about me I have a LinkedIn profile at http://www.linkedin.com/in/sylvainbellemare. *DOCUMENTATION* Documentation-wise, I can use LaTeX. For instance, my resume, which is attached with this email, was written with LaTeX. It's noteworthy to mention that I like to work on documentation, as long as it balances with coding work [?]. If that makes any sense to you and you have ideas on projects that I could contribute to, please let me know, and I'll be very happy to do my very best to contribute. Best regards, -Sylvain -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: 328.png Type: image/png Size: 569 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: sylvain.bellemare.pdf Type: application/pdf Size: 134849 bytes Desc: not available URL: