Mailman 3 February 2010 - yt-dev

Re: [Yt-dev] minor error in units
by Matthew Turk 05 Feb '10

05 Feb '10

Hi JC, You're totally right. I think when I wrote it, I thought I was taking the diameter of the sun and dividing by two, but I was really taking the radius of the sun and dividing by two. You are reporting a bug that has been in yt since the very first day I worked on it. Yowza. Thanks, Matt On Fri, Feb 5, 2010 at 1:35 PM, Jean-Claude Passy <jcpassy(a)gmail.com> wrote: > Hi Matt, > > just to let you know. I think I have found a very minor bug in units > conversion: > > > ######################################################### > In [2]: filen = 'static_nogas0000' > > In [4]: import yt.lagos as L > > In [6]: data.units > Out[6]: > {'1': 1, > 'au': 3.4409335988416996, > 'aye': 1.0, > 'cm': 51467226398186.297, > 'kpc': 1.6679270958999999e-08, > 'miles': 319741624.28402996, > 'mpc': 1.6679270958999998e-11, > 'pc': 1.6679270958999999e-05, > 'rsun': 369.72939934815298, > 'unitary': 1.0} > > In [7]: 51467226398186.297 / 6.955e10 > Out[7]: 740.00325518599993 > > In [8]: 51467226398186.297 / 1.496e13 > Out[8]: 3.4403226201996189 > > In [9]: 51467226398186.297 / 3.0857e21 > Out[9]: 1.6679270958999999e-08 > ######################################################### > > There is an error in the filed 'rsun'. It should give around 740, not 370. > The rest looks fine. > > Take care, > > JC >

1 0

numpy 1.4
by Stephen Skory 03 Feb '10

03 Feb '10

Hi all, I am using numpy.in1d in the merger tree, which is new in numpy 1.4. http://docs.scipy.org/doc/numpy/reference/generated/numpy.in1d.html Do we think that this will be a problem, kinda sorta forcing users to upgrade when we do the next point release? I think as long as we say 'numpy 1.4 required' it's ok. _______________________________________________________ sskory(a)physics.ucsd.edu o__ Stephen Skory http://physics.ucsd.edu/~sskory/ _.>/ _Graduate Student ________________________________(_)_\(_)_______________

2 1

yt docs
by Stephen Skory 03 Feb '10

03 Feb '10

Hi all, just so all of you know, I've started docs on the Merger Tree, and since that will be 'released' the next point release, I've created a yt-1.7 branch in the docs repo. That is all for this email. _______________________________________________________ sskory(a)physics.ucsd.edu o__ Stephen Skory http://physics.ucsd.edu/~sskory/ _.>/ _Graduate Student ________________________________(_)_\(_)_______________

1 0

Re: [Yt-dev] Halo Merger Tree
by Stephen Skory 01 Feb '10

01 Feb '10

Eric, > 1. Have you run this on the light cone simulation yet? No. If that is to happen, I'd like this tool to be shaken down a bit more before the resources are applied to that. But that's a good idea! > 2. I don't know how to query an SQLite database, can you help me out with that? Sure. The primary reason why I'm using a SQL database is all the big catalogs of observational (SDSS) and simulated data (Millennium) make the data publicly accessible via a SQL interface. What I've written isn't quite up to the scope of those databases, but as it is improved it can be (& more!). There are lots of examples of how to write SQL queries out there, but I'll write some examples specific to this database soon for you. > 3. It looks like you are appending filenames to your list in the example below, > have you thought about using the EnzoSimulation class to call it and run it? I haven't used that class before so I didn't even think to use it. I'll look into that, thanks for reminding me! > 4. I also don't know about GraphViz, maybe you could help me out with that as > well. Graphviz is a open-source tool that can build visualizations of relational data. It's not as pretty as you could do by hand with a professional graphics tool, but it has an easy to output flexible markup language and that's why I use it. There are Graphviz-Python packages, but the ones I can find have at least two years since last update, and would require someone to install on their Python, which most people have a hard time doing. Therefore I hard-coded the Graphviz output content into the source. If you want to change the output style and content, you can use what I wrote as a starting point. http://graphviz.org/ _______________________________________________________ sskory(a)physics.ucsd.edu o__ Stephen Skory http://physics.ucsd.edu/~sskory/ _.>/ _Graduate Student ________________________________(_)_\(_)_______________

2 4