Mailman 3 July 2016 - yt-users

Clump surface pressure
by John Regan 21 Jul '16

21 Jul '16

Hi All, I was hoping to calculate the surface pressure on the surface of a clump. I see that there is a way to do this with surfaces at the moment in YT probably by doing something like: mysurf.calculate_isocontour_flux(*"density"*, *some_value*, *r_x*, *r_y*, *r_z*, *"pressure"*) So while that has a fair chance of working for a surface? Is there any way to get a clump to do the same? Any tips/advice greatly appreciated! Cheers, John

1 0

yt parallelism with mpi4py at OLCF
by Max Katz 17 Jul '16

17 Jul '16

Hi all, I'm trying to get yt parallelism enabled to run on the visualization cluster at OLCF. I am using a conda environment for the yt prerequisites but I have the development branch of yt installed from source. To get the parallelism working I tried installing mpi4py both using pip and building it from source (with the latter, explicitly telling it that the MPI-enabled C compiler is the wrapper cc). However, the statement "from mpi4py import MPI" that yt uses to initialize the session fails with: >>> from mpi4py import MPI [Sun Jul 17 03:52:10 2016] [unknown] Fatal error in PMPI_Init_thread: Other MPI error, error stack: MPIR_Init_thread(547): MPID_Init(203).......: channel initialization failed MPID_Init(584).......: PMI2 init failed: 1 Aborted Some googling suggests that this is potentially an issue specific to the way Cray sets up their MPI installation. At any rate, although this is not necessarily a yt-specific issue, if anyone has relevant experience or knows a workaround, please let me know, thanks! Max Katz Ph.D. Candidate, Department of Physics and Astronomy Stony Brook University http://astro.sunysb.edu/mkatz/

1 0

Re: [yt-users] Using AMR with inline python
by Azton Wells 14 Jul '16

14 Jul '16

Thanks Matt for your reply. I am using rho = na.asarray(enzo.grid_data[enzo.grid_data.keys()[0]]["Density"]) to access the density, and then insert values into another field, so I'm not using them for plotting or anything. So after checking the density of a cell and setting the value 'emis' accordingly, I use enzo.grid_data[enzo.grid_data.keys()[0]][field][cellx][celly][cellz]=enzo.grid_data[enzo.grid_data.keys()[0]][field][cellx][celly][cellz]+emis to modify a field within the simulation. Unfortunately, the value 'emis' that I am generating is volume dependent, so I need a method to find the volume of the cell (in a way consistent with the two methods above, if possible). Failing finding the volume, if I could find the AMR level, I would automatically know the volume. Is this possible using the enzo object? Sorry if this is an obvious one, I am relatively new to enzo & yt. Best, Azton

1 0

origin in SlicePlot
by Iapichino, Luigi 14 Jul '16

14 Jul '16

Dear yt users, I have a question on the "origin" parameter in SlicePlot: is it possible to have the location of the origin of the plot coordinate system in a user-defined point inside the slice? I am plotting a galaxy cluster where, for sake of visualisation, the center is neither in the slice center, nor in one of the corners, and I would like the coordinate origin there. It looks like something possible only with a lot of efforts, but I hope I am wrong! Looking forward to your suggestions, Best regards, Luigi --- Dr. Luigi Iapichino Leibniz-Rechenzentrum Application Support group Boltzmannstr. 1, 85748 Garching, Germany Tel.: +49 89 35831 7824 e-mail: luigi.iapichino(a)lrz.de

2 2

lightray start and end points
by Stephanie Tonnesen 13 Jul '16

13 Jul '16

Hi yt-users, I have a somewhat strange goal using lightray. I have a large box, 120 Mpc/h on a side, but only a portion of it is highly resolved (about 25 Mpc/h on a side). I want to create a set of lightrays that span a small, constant z range and only cross the refined region of the box. So right now I am using the cosmology LightRay call so that I can set the redshift range, and would like to be able to take a look at the start and stop positions of the lightray that is created, and not just in the output dump of the run. What command can I use to make those positions into a variable in my code? I tried lr.start_position, but that doesn't exist. Any help is appreciated! Thanks, Stephanie -- Dr. Stephanie Tonnesen Alvin E. Nashman Postdoctoral Fellow Carnegie Observatories, Pasadena, CA stonnes(a)gmail.com

3 2

Greeting and questions.
by Martín Mestre 12 Jul '16

12 Jul '16

Dear yt community, a week ago I knew about this project and after reading a small part of the manual, the mission statement and the code of conduct I got encouraged to try to learn yt (and python). I would be very glad if I could use yt to study the structure of some disrupted satellites that I evolved with Gadget-2 (dark matter only). I have installed yt and made a simple halo catalog but I am not sure if I have proceeded correctly. I have written some questions below. Thank you very much in advance for any answer or suggestion. Best regards from La Plata, Argentina, Martín Mestre ---------------- INSTALLATION The Host information is: Linux seminare 2.6.37.1-mosix #2 SMP Tue Apr 19 12:57:51 ART 2011 x86_64 GNU/Linux. I installed yt 3.2.3 as a simple cluster user in my home directory with the installation script. I have run the script a second time setting the option: INST_ROCKSTAR=1. The installation log is too large for the pastebin. I have only pasted the lines with the words "error" and "Error" respectively in the links: http://paste.yt-project.org/show/6657/ and http://paste.yt-project.org/show/6658/ . The first/last 400 lines of the log are here: http://paste.yt-project.org/show/6659/ and http://paste.yt-project.org/show/6660/ . How could I proceed to eliminate these errors ? ----------------------------------------------------------- LOADING GADGET-2 SNAPSHOT IN BINARY FORMAT AND COMPUTATION OF HALO CATALOG I have run the following scritp: http://paste.yt-project.org/show/6661/ and the log files was: http://paste.yt-project.org/show/6662/ I would like to know if everything looks right in the log file. I wasn't able to upload the image file (36KB), saying: ERROR HTTP Error 404: Not Found I will be willing to send it by email because I have a question to do about the use of the halo finder. It found 8 halos, one of them many orders of magnitude more massive than the rest. This was expected because it identifies the stream progenitor. The location of this halo is much closer to the centre of mass of the whole ensemble than to the point of maximum density (which I had previusly found through the method of the "shrinking spheres"). Is there any scale parameter of the halo finder that allows me to identify halo centres as the points of local maxima of the volumetric density ? ---------------------------------------------------------- ---------------------------------------------------------- .....

3 6

a possible teensy typo?
by Stephanie Tonnesen 09 Jul '16

09 Jul '16

When I install yt-dev and then run the install_script.sh it tells me that list of what it will and won't be installing. The first thing on the list is: INST_CONDA = 1 so I won't be installing a conda-based python environment It uses this miniconda python environment, and all the rest of the 1's mean "will" instead of "won't". Is this a typo or do I just not understand what it is telling me? Thanks, Stephanie -- Dr. Stephanie Tonnesen Alvin E. Nashman Postdoctoral Fellow Carnegie Observatories, Pasadena, CA stonnes(a)gmail.com

2 1

Set colors of bad values for a colormap?
by Yi-Hao Chen 08 Jul '16

08 Jul '16

Hi all, I have some projection plots that are displayed in log scale. However, some of the pixels have "0" values. The log10 of zero is -inf and yt displays them in white. I am looking for a way to set the color of those pixels. Does anyone know how to do it? Something similar to set_bad() <http://matplotlib.org/api/colors_api.html#matplotlib.colors.Colormap.set_bad> in matplotlib would be really helpful. Cheers, Yi-Hao Chen

2 2

Loading dataset with no particles using a script with particle fields
by José Utreras 07 Jul '16

07 Jul '16

Hi yt users, I'm using a script to calculate things for cells and particles when needed. But I'm having the following problem: - Each time I try to work with datasets that do not contain particles the particle fields are loaded which results in an error. In my simulations some datasets have particles at the beginning and other at later times. Then it is not desirable to have different scripts for datasets with and without particles. Is there a way to tell yt to not load field particles when particles are not present? Best regards Jose

2 1

Creating Field to calculate enclosed mass
by Rajika Kuruwita 06 Jul '16

06 Jul '16

Hi yt users, I've been using yt for a few years, and I know how to simple fields, but I'm wondering if someone has created a field that calculates enclosed mass at all cells (so, I guess the mass in a sphere centred on the centre of the simulation domain, or centre of mass). I've tried to think of a few ways of doing this, but all require iterating over all cells, and that's not very efficient. If anyone has an suggestions or can help, that would be great! Regards, Rajika.

2 1