Hi all,
I've added code to yt.extensions.StarAnalysis (in hg, of course) to calculate the spectral flux and spectral energy distribution for a collection of stars. This is based on code from Ken Nagamine, and it uses the data tables provided by Bruzual & Charlot. You can pass it a data_source, or arrays containing the pertinent information (star mass, creation time and metallicity) and it will build you a spectrum.
The rub is in the B&C data tables. They are distributed as ASCII files from an unreliable foreign server. The data required to build a spectrum is at least 150MB in ASCII format and takes a looong time to download. And then once you have it, reading ASCII data in is super sloooow. I have converted the data to HDF5 files which are nice and fast and only 72MB in size in total. However, I'm not sure exactly if we can distribute them as HDF5 files. I am going to email B&C and ask them if we can, with proper acknowledgement, of course. Otherwise, I'll have to include my (currently ridiculously slow) conversion script with any instructions on how to use the spectrum code.
I'll let you guys know!
sskory@physics.ucsd.edu o Stephen Skory http://physics.ucsd.edu/~sskory/... _.>/ _Graduate Student __()\()__
All,
The rub is in the B&C data tables.
Good news! I emailed them and Gustavo Bruzual responded promptly giving me his permission. Also, he seems genuinely curious about HDF5 and I gave him a quick overview of it, and a link to the main HDF5 webpage. He said they are readying a new set of tables and I'm guessing he wants a better way to package the data than they did last time.
Let me know if you have any questions (as in you wish to use the code, as there is no docs for it yet).
sskory@physics.ucsd.edu o Stephen Skory http://physics.ucsd.edu/~sskory/... _.>/ _Graduate Student __()\()__
Hi Stephen,
Good news! I emailed them and Gustavo Bruzual responded promptly giving me his permission. Also, he seems genuinely curious about HDF5 and I gave him a quick overview of it, and a link to the main HDF5 webpage. He said they are readying a new set of tables and I'm guessing he wants a better way to package the data than they did last time.
This is GREAT news. Thank you very much for taking the lead on this, and for adding this to the code. I think there's some great science to be done with this, and I like the direction you are taking with the code.
Again, good work, and thanks.
-Matt
Good news! I emailed them and Gustavo Bruzual responded promptly giving me his permission. Also, he seems genuinely curious about HDF5 and I gave him a quick overview of it, and a link to the main HDF5 webpage. He said they are readying a new set of tables and I'm guessing he wants a better way to package the data than they did last time.
>
This is GREAT news. Thank you very much for taking the lead on this, and for adding this to the code. I think there's some great science to be done with this, and I like the direction you are taking with the code.
Again, good work, and thanks.
-Matt
I'll second that. This is a super-useful thing to have in yt, and I'm sure we'll get a lot of mileage out of it. Are you going to post the hdf5 file on a website somewhere, then?
--Brian
Brian,
>
I'll second that. This is a super-useful thing to have in yt, and I'm sure we'll get a lot of mileage out of it. Are you going to post the hdf5 file on a website somewhere, then?
Sure. I'm going to write up some docs and then have Matt upload the files to his server and send another email. I'll try to do that by tonight, just to git 'er done. It isn't terribly difficult to use (which is the point of yt-izing it) but I should document my work!
sskory@physics.ucsd.edu o Stephen Skory http://physics.ucsd.edu/~sskory/... _.>/ _Graduate Student __()\()__
Brian,
I'll second that. This is a super-useful thing to have in yt, and I'm sure we'll get a lot of mileage out of it. Are you going to post the hdf5 file on a website somewhere, then?
Matt uploaded them here:
http://yt.enzotools.org/files/bc03/...
I've added instructions to the yt-doc repo:
http://hg.enzotools.org/yt-doc/file/e449707b6046/source/extensions/star_anal...
sskory@physics.ucsd.edu o Stephen Skory http://physics.ucsd.edu/~sskory/... _.>/ _Graduate Student __()\()__
As a quick note, I'm going to rebuild the docs after I update them to the current state of the trunk, and then deploy a new build.
These will be for the 1.6 release, which (being a point release) is almost all done anyway and will be pushed out by the New Year.
On Tue, Dec 15, 2009 at 7:05 PM, Stephen Skory stephenskory@yahoo.com... wrote:
Brian,
I'll second that. This is a super-useful thing to have in yt, and I'm sure we'll get a lot of mileage out of it. Are you going to post the hdf5 file on a website somewhere, then?
Matt uploaded them here:
http://yt.enzotools.org/files/bc03/...
I've added instructions to the yt-doc repo:
http://hg.enzotools.org/yt-doc/file/e449707b6046/source/extensions/star_anal...
___ sskory@physics.ucsd.edu o Stephen Skory http://physics.ucsd.edu/~sskory/... _.>/ _Graduate Student __()()__
Yt-dev mailing list Yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org...
On Tue, Dec 15, 2009 at 10:05 PM, Stephen Skory stephenskory@yahoo.com...wrote:
Brian,
I'll second that. This is a super-useful thing to have in yt, and I'm sure we'll get a lot of mileage out of it. Are you going to post the hdf5 file on a website somewhere, then?
Matt uploaded them here:
http://yt.enzotools.org/files/bc03/...
I've added instructions to the yt-doc repo:
http://hg.enzotools.org/yt-doc/file/e449707b6046/source/extensions/star_anal...
Very cool. Thanks!
Brian
-
Brian O'Shea -
Matthew Turk -
Stephen Skory