Mailman 3 February 2014 - yt-dev

Big performance boost on unitrefactor.
by Nathan Goldbaum 14 Feb '14

14 Feb '14

Hi all, If you've been using unitrefactor, updating to the latest version should come with a big performance boost for nearly all yt operations. More details here: https://bitbucket.org/MatthewTurk/yt/pull-request/67/refactor-the-unit-syst… To update: $ hg pull -r yt-3.0 https://bitbucket.org/MatthewTurk/yt $ hg up $ python setup.py develop Be sure to save any uncommitted changes with hg shelve first. -Nathan

1 0

new features in development: halo analysis
by Britton Smith 14 Feb '14

14 Feb '14

Hi all, In case you are not getting the notifications for PRs to the YTEPs, I wanted to draw your attention to an update I just made to YTEP 0012, which proposes a new halo analysis framework to be built in yt-3.0. You can view that here: https://bitbucket.org/yt_analysis/ytep/pull-request/33/updating-ytep-0012-h… A good amount of work has been done on what will be a complete redesign of how we analyze halos en masse with yt. In other words, this will take the place of the HaloProfiler and become something much greater. The YTEP details most of this, but if you wish to check out the source, it is in my fork (https://bitbucket.org/brittonsmith/yt/) in the "ytep0012" bookmark. The framework is moderately functional at the moment as I am developing it in support of my current astro research. That said, since I am developing with an eye toward my own work, I might not have time to flesh out functionality that I currently don't need for a while. If anyone is interested in using or working on this, please do get involved. For now, the easiest way will probably be to fork my fork and issue PRs. One additional note, this work is being done within the new unit refactor, so some other infrastructure things may appear quite different to those unfamiliar with that work. Anyway, please don't hesitate to contact me or this list if you want to get involved. Any and all help/feedback is welcome! Britton

1 0

kdtree tiles and pf.h.grids
by Alex Bogert 14 Feb '14

14 Feb '14

Hi, Matt asked me to write yt-dev about yt-3.0 kdtree decomposition tutorials. I'm looking for a robust data flow to do volume rendering of AMR data on the GPU. >From our email chain : " But the grids can just be accessed by getting pf.h.grids. To get blocks from the tree in 3.0 we have a .tiles attribute on the data source which will do the full decomposition into non overlapping tiles. If you write to yt-dev this will also help me remember to document this." -mjturk Alex

2 1

Issue #782: amr kd-tree doesn't work with units refactor (yt_analysis/yt)
by Matthew Turk 14 Feb '14

14 Feb '14

New issue 782: amr kd-tree doesn't work with units refactor https://bitbucket.org/yt_analysis/yt/issue/782/amr-kd-tree-doesnt-work-with… Matthew Turk: This script fails because of unit conversion issues. ``` #!python from yt.mods import * pf = load("HiresIsolatedGalaxy/DD0044/DD0044") dd = pf.h.all_data() tiles = dd.tiles for grid, node, (sl, dims, gi) in tiles.slice_traverse(): print grid ``` Responsible: samskillman

1 0

IPython 2.0 widgets
by Nathan Goldbaum 14 Feb '14

14 Feb '14

Hi all, I want to share a cool notebook I just made with you all. It uses an IPython 2.0 javascript widget to interactively control a yt SlicePlot. Since the widget needs to talk to a live kernel, you'll need to run the notebook locally to get the full awesomeness. http://nbviewer.ipython.org/gist/ngoldbaum/8971185 This took me all of ten minutes to set up based on Fernando's example [1] and works pretty much out of the box with IPython 2.0. I highly encourage you to clone the IPython repo and test out the new version on the master branch. I think this will open up a lot of room for interactivity and discovery with yt's plot objects. -Nathan [1] https://twitter.com/fperez_org/status/433856794436194304

4 6

yt-project.org down?
by Nathan Goldbaum 13 Feb '14

13 Feb '14

Hi all, It looks like yt-project.org is down. Just a heads up in case there are some cages that need to be rattled at dreamhost. -Nathan

2 1

Issue #781: Deposited particle fields in Enzo not working (yt_analysis/yt)
by John Wise 12 Feb '14

12 Feb '14

New issue 781: Deposited particle fields in Enzo not working https://bitbucket.org/yt_analysis/yt/issue/781/deposited-particle-fields-in… John Wise: yt-3.0 cannot generate deposited fields in Enzo datasets. For example, when I'm trying to access either 'dm_density', ['deposit', 'all_density'], or ['cic_particle_velocity_x'] in a single grid with the following lines ``` #!python pf = load(sys.argv[-1]) g = pf.h.grids[0] dm = g['dm_density'] ``` It crashes with the following error, ``` #!python yt.utilities.exceptions.YTCouldNotGenerateField: Could field '('all', 'particle_position_x')' in output_0000 could not be generated. ``` even though it's in the dataset, and executing g['all', 'particle_position_x'] gives no errors. It looks like _generate_particle_field() is returning None.

1 0

Enzo derived particle fields in 3.0
by John Wise 12 Feb '14

12 Feb '14

Hi everyone, I've run into the problem where yt-3.0 cannot generate deposited fields in Enzo datasets. For example, when I'm trying to access either 'dm_density', ['deposit', 'all_density'], or ['cic_particle_velocity_x'] in a single grid with the following lines pf = load(sys.argv[-1]) g = pf.h.grids[0] dm = g['dm_density'] It crashes with the following error, yt.utilities.exceptions.YTCouldNotGenerateField: Could field '('all', 'particle_position_x')' in output_0000 could not be generated. even though it's in the dataset, and executing g['all', 'particle_position_x'] gives no errors. It looks like _generate_particle_field() is returning None, and I have no idea why since I'm not very familiar with the new field objects in 3.0. Does anyone know what's going on? Thanks, John -- John Wise Assistant Professor of Physics Center for Relativistic Astrophysics, Georgia Tech http://cosmo.gatech.edu

2 1

Issue #780: particle unions getting lost in TimeSeries Enzo data (yt_analysis/yt)
by John ZuHone 12 Feb '14

12 Feb '14

New issue 780: particle unions getting lost in TimeSeries Enzo data https://bitbucket.org/yt_analysis/yt/issue/780/particle-unions-getting-lost… John ZuHone: The particle union "all" is getting lost from pf.particle_types when you access it from a TimeSeries, making it impossible to get certain fields. Example notebook: https://hub.yt-project.org/nb/oduwuj Responsible: MatthewTurk

1 0

Changing log level from within yt
by John ZuHone 12 Feb '14

12 Feb '14

Hi all, I have an analysis module that loops over a TimeSeries object. I'd like to temporarily reduce the loglevel so that we don't see all of the stuff from loading each parameter file, and then re-set it when I'm done. I tried this before the loop: ytcfg["yt", "loglevel"] = "40" which basically does what I need it to, while still keeping the progress bar around, and then after the loop I tried setting the loglevel back to its previous value. However, this doesn't work. It only seems to work if I set the loglevel before calling yt.mods which calls the analysis module. Is there a way to temporary suppress logging info? Best, John

2 1