Mailman 3 August 2015 - yt-users

getting slice array
by Slavin, Jonathan 11 Aug '15

11 Aug '15

Hi all, I'm new to using yt, though I'm an experienced python and matplotlib user. I've been doing runs with FLASH and would like to be able to plot the results with matplotlib. I've used yt "interactively" in an ipython notebook and found it a bit cumbersome - e.g. you can't pan and zoom like you can with matplotlib. On the other hand yt has some nice facilities for accessing the data. So my question is, how do I get a slice, such as is plotted using the yt.SlicePlot function, in an array that I can then manipulate, plot, etc.? If I do: ds = yt.load(file) slc = ds.slice(2,0.) d = slc['density'] I have a YTArray that's apparently 1-D: d.shape (138496,) I should mention that this is a 2-D, cylindrically symmetric (r-z) run. Thanks in advance for any help you can offer. Regards, Jon Slavin -- ________________________________________________________ Jonathan D. Slavin Harvard-Smithsonian CfA jslavin(a)cfa.harvard.edu 60 Garden Street, MS 83 phone: (617) 496-7981 Cambridge, MA 02138-1516 cell: (781) 363-0035 USA ________________________________________________________

3 2

Re: [yt-users] getting slice array
by Slavin, Jonathan 11 Aug '15

11 Aug '15

Hi Nathan, Thanks! It seems though that the frb cannot be directly passed to matplotlib's imshow, though np.array(frb) can. I would like it in general if yt could provide simpler means of accessing arrays or array-like objects such as are plotted up. There are many reasons for this besides the desire to plot using matplotlib -- e.g. to create plots of ratios of quantities, or to plot scaled variables. Perhaps this is possible now, though it's not obvious from the docs. Regards, Jon On Tue, Aug 11, 2015 at 2:42 PM, <yt-users-request(a)lists.spacepope.org> wrote: > Date: Tue, 11 Aug 2015 13:17:50 -0500 > From: Nathan Goldbaum <nathan12343(a)gmail.com> > To: Discussion of the yt analysis package > <yt-users(a)lists.spacepope.org> > Subject: Re: [yt-users] getting slice array > Message-ID: > <CAJXewOmiymap0uUBy0hL= > zo0Ltf0kG7UErXYFy9dqZseXoapsg(a)mail.gmail.com> > Content-Type: text/plain; charset="utf-8" > > On Tue, Aug 11, 2015 at 1:10 PM, Slavin, Jonathan <jslavin(a)cfa.harvard.edu > > > wrote: > > > Hi all, > > > > I'm new to using yt, though I'm an experienced python and matplotlib > > user. I've been doing runs with FLASH and would like to be able to plot > > the results with matplotlib. I've used yt "interactively" in an ipython > > notebook and found it a bit cumbersome - e.g. you can't pan and zoom like > > you can with matplotlib. On the other hand yt has some nice facilities > for > > accessing the data. So my question is, how do I get a slice, such as is > > plotted using the yt.SlicePlot function, in an array that I can then > > manipulate, plot, etc.? If I do: > > ds = yt.load(file) > > slc = ds.slice(2,0.) > > d = slc['density'] > > I have a YTArray that's apparently 1-D: > > d.shape > > (138496,) > > I should mention that this is a 2-D, cylindrically symmetric (r-z) run. > > Thanks in advance for any help you can offer. > > > > If you have a yt plot object (e.g. a SlicePlot or a ProjectionPlot), you > can do: > > slc = yt.SlicePlot(...) > densit_image = slc.frb['density'] > > "frb" here is a FixedResolutionBuffer object which translates the > multiresolution 1-D array you ran into above into a 2D pixelized > representation of your data. > > You can also create a FixedResolutionBuffer object directly: > > http://yt-project.org/docs/dev/analyzing/generating_processed_data.html#d-i… > > The image arrays you get back from a FixedResolutionBuffer object can be > passed directly to e.g. matplotlib's imshow command. > > By the way, Matt Turk has an open pull request to add the interactive > panning and zooming you were looking for: > > > http://yt-project.org/docs/dev/analyzing/generating_processed_data.html#d-i… > > I'm hoping to finish up that pull request soon, since having interactive > plots both in the notebook and using matplotlib's interactive backends is a > common feature request. > -- ________________________________________________________ Jonathan D. Slavin Harvard-Smithsonian CfA jslavin(a)cfa.harvard.edu 60 Garden Street, MS 83 phone: (617) 496-7981 Cambridge, MA 02138-1516 cell: (781) 363-0035 USA ________________________________________________________

2 1

volume rendering colorbar issue
by Stephanie Tonnesen 11 Aug '15

11 Aug '15

Hi yt-users, I am trying to make a volume rendering using --- Version = 3.2-dev Changeset = dc2467c4eae7 (yt) tip --- and basically following this example: http://bender.astro.sunysb.edu/random/vol_rotate.py My problem is that the units are wonky on my colorbar. I input them in my transfer function as cgs, and the peaks seem to be in the right place based on my input range and sample values. Can I get the colorbar to present the density values in cgs? You can see the output and the code attached. Oh, I had to add in the line matplotlib.use('pdf') because I was getting a _tkinter can't find your screen error everytime I tried to add the colorbar. I think it was a matplotlib with my computer problem. Thanks! Stephanie -- Dr. Stephanie Tonnesen Alvin E. Nashman Postdoctoral Fellow Carnegie Observatories, Pasadena, CA stonnes(a)gmail.com

1 0

problems installing rockstar with yt+anaconda
by Brian O'Shea 11 Aug '15

11 Aug '15

Hi folks, I'm using Enzo on Blue Waters, and to get it working I needed to install anaconda and then install yt-3 from source (i.e., by 'hg clone https://bitbucket.org/yt_analysis/yt' and then 'python setup.py develop'). This does not install rockstar by default, unfortunately, and I am having trouble getting rockstar working (although yt itself works great). I've tried to essentially mimic the rockstar portion of the install script for yt-3 (https://bitbucket.org/yt_analysis/yt/raw/yt/doc/install_script.sh) by doing the following: > hg clone http://bitbucket.org/MatthewTurk/rockstar > hg pull > hg up -C tip > make lib > cp librockstar.so ~/local_lib/lib/. and then pointing to ~/local_lib/lib with LD_LIBRARY_PATH (using the full path, not the '~/' convention). However, even when I do this I still can't seem to persuade the rockstar interface code in yt to find rockstar. Attempting to import rockstar in a script by: from yt.analysis_modules.halo_finding.rockstar.api import RockstarHaloFinder results in this error: ImportError : from .rockstar import RockstarHaloFinder cannot import name rockstar_interface and the script immediately exits. When I look at the directory yt/analysis_modules/halo_finding/rockstar in a version of yt that has been installed from a script, there are some additional .so files (rockstar_groupies.so and rockstar_interface.so) that aren't in my source install, even when I've run 'python setup.py develop' in that directory. I feel like I'm doing something obviously wrong here. Any suggestions? Thanks! --Brian

2 2

annotate_contour takes much longer time to run than expected on some machines
by Bili Dong 11 Aug '15

11 Aug '15

Hi yt-users, I’m using yt-3.2 (Changeset = 329b5039b031 stable). The problem is that the same script that could finish in seconds on my local machine will take much (order of magnitude) longer time to run on the server I’m using. In some cases, it get stuck for hours that I don’t know if it will ever finish. I’ve tracked down the bottleneck to be the `annotate_contour` method. The server should be faster than my local machine. (I’m not exactly sure about how to compare it. But the server is generally much faster in other situations.) For example, for this script (http://paste.yt-project.org/show/5808/ <http://paste.yt-project.org/show/5808/>), on my local machine the output shows: ``` yt : [INFO ] 2015-08-10 14:48:15,401 Making a fixed resolution buffer of (('gas', 'density')) 800 by 800 yt : [INFO ] 2015-08-10 14:48:15,751 Saving plot snapshot_200_Projection_z_density.png yt : [INFO ] 2015-08-10 14:48:16,716 Saving plot snapshot_200_Projection_z_density.png ``` However, on the server the output shows: ``` yt : [INFO ] 2015-08-10 14:51:25,551 Making a fixed resolution buffer of (('gas', 'density')) 800 by 800 yt : [INFO ] 2015-08-10 14:51:25,917 Saving plot snapshot_200_Projection_z_density.png yt : [INFO ] 2015-08-10 14:56:21,922 Saving plot snapshot_200_Projection_z_density.png ``` That’s 296 times slower!!! This is likely to be a very specific problem with the server. But I’d like to know if anyone else has encountered similar problems before so I can learn from your experience. For now, I have no idea what could have gone wrong. BTW, I’m using TSCC (http://idi.ucsd.edu/computing/ <http://idi.ucsd.edu/computing/>) at UC San Diego, if you happen to know it. Thanks, Bili

5 10

custom axes tickmarks projection plot
by Gabriel Goodwin 10 Aug '15

10 Aug '15

Hello All, Does know how to specify custom tickmark locations for the y axis of a projection plot in the 3.2 release? I was able to do it in 3.1 using the following code: p = yt.ProjectionPlot(ds,2,string,center=[cent,height,0],width=(wide,height),origin="native",data_source=cut,method='integrate',weight_field='temperature') ax = p.plots[string].axes ax.yaxis.set_ticks([0.00, 0.15, 0.30]) This would place tickmarks and their labels on the y axis at 0, 0.15, and 0.3. When I try and do this using yt 3.2 I get the following error: AttributeError Traceback (most recent call last)<ipython-input-76-6e901a613181> in <module>() 16 ax = p.plots[string].axes---> 17 ax.yaxis.set_ticks([0.00, 0.15, 0.30]) AttributeError: 'NoneType' object has no attribute 'yaxis' It appears that we can no longer manipulate the axes object like we were able to in 3.1. Thank you!

2 2

annotate_grids() not working (Enzo)
by bh11e 10 Aug '15

10 Aug '15

Hello, yt users! I visualize Enzo output with yt (version = 3.1-dev, Changeset = f6379bbc462b+ (yt) tip). Sometimes annotate_grids() doesn't work neither for slice plots nor for projection plots. This affects all data sets from a given simulation. At the same time other annotations like annotate_line() display correctly. Had anyone gad a similar problem before? Many thanks, --Boyan Hristov

2 1

data_source=region
by Shankar Agarwal 06 Aug '15

06 Aug '15

Hi, http://yt-project.org/doc/cookbook/complex_plots.html#thin-slice-projections Is it possible to create a thin slice projection of particles (instead of density)? I tried the following three methods but get errors: 1. p = yt.ParticlePlot(ds, 2, ['particle_mass'], data_source=region) TypeError: __init__() got multiple values for keyword argument 'data_source' 2. p = yt.ParticleProjectionPlot(ds, 2, ['particle_mass'], data_source=region) finfo = self.ds._get_field_info("unknown", fname) 3. p = yt.ProjectionPlot(ds, 2, ['particle_mass'], data_source=region) File "/home/sagarwal/codes/yt/yt-x86_64/src/yt-hg/yt/visualization/plot_window.py", line 1429, in __init__ field_parameters = field_parameters, method = method) File "/home/sagarwal/codes/yt/yt-x86_64/src/yt-hg/yt/data_objects/construction_data_containers.py", line 253, in __init__ field = self._determine_fields(ensure_list(field)) File "/home/sagarwal/codes/yt/yt-x86_64/src/yt-hg/yt/data_objects/data_containers.py", line 518, in _determine_fields finfo = self.ds._get_field_info("unknown", fname) Shankar

2 2

Custom Colorbar Axis
by Gabriel Goodwin 05 Aug '15

05 Aug '15

Hello All, Does anyone know how to specify the tickmark locations on the colorbar axis? I am able to do this on the plot axes without issue, but have not been able to successfully do the same with the colorbar. Attached is a pic of one of my plots using default colorbar axis ticks. I have tried to use the following code (and many variants of it I have found in the python/matplotlib documentation) to specify the location of the tickmarks, but they don't place the ticks accurately. My data is temperature plotted over the range of 300K to 5500K and I want to have tickmarks at 300K, 800K, 3500K, and 4500K. Also attached is pic of my plot showing an attempted custom colorbar axis. I'm not sure if it makes a difference, but I made the colormap in matplotlib. colorbar_axes = p.plots[string].cax colorbar_axes.yaxis.set_ticks([0, 0.1, 0.8]) Thanks for the help!

2 1

Interesting crash when using ("deposit", "all_cic")
by Brian O'Shea 04 Aug '15

04 Aug '15

Hi folks, I have been trying to make some dark matter density movies using the "all_cic" field, and have discovered that using "all_cic" can reliably cause yt to seg fault on the supercomputer I'm using (x86_64 linux cluster with kernel 2.6.32-504.8.1.el6.x86_64) using the tip of yt-dev (changeset fa08e386d0da). The script that causes the crash is here: http://paste.yt-project.org/show/5774/ After I turned up the log level and started using pdb to debug, I discovered that the line in the script that causes the seg fault is line 88: dm_dens_x = my_reg[("deposit", "all_cic")].value where my_reg is defined as (on line 82): my_reg = ds.arbitrary_grid(left, right, dims=[1,800, 800]) After stepping through with pdb, it seems that the code dies in this function: /mnt/home/oshea/yt-3/lib/python2.7/site-packages/sympy/core/basic.py(83)__new__() with the traceback shown here: http://paste.yt-project.org/show/5775/ Based on this traceback, it seems that there are problems with the unit system. Interestingly enough, though, this does not happen on every dataset - it seems to happen occasionally, but predictably. For example, I can generate two images using this script, but it will seg fault on the third. On a different dataset in the time series, it will produce one image and then seg fault on the second. If I run the script over and over, it will eventually produce quite a few images, until it runs into a situation where it seg faults on the first image I produce (at which point I went to the debugger). I don't have enough experience with the guts of yt - and the units section of yt in particular - to have a sense of what might be happening here based on the traceback. Does anybody have any suggestions? Thank you! --Brian

2 3

overplotting halos gives different results ??
by Shankar Agarwal 04 Aug '15

04 Aug '15

Hi, When I try to overplot halos I see that some of the circles go missing, depending on what "center" is use for the density projection. The following two attempts gave me different number of circles. Cheers, Shankar ______________________________________________________________ Try 1: sim_ds = yt.load('Enzo_64/RD0006/RedshiftOutput0006') halos_ds = yt.load('halo_catalog/halo_catalog.0.h5') hc = HaloCatalog(data_ds=sim_ds, halos_ds=halos_ds) hc.load() p = yt.ProjectionPlot(sim_ds, "x", "density", center=(0.5,0.5,0.5)) p.annotate_halos(hc) ______________________________________________________________ ______________________________________________________________ Try 2: p = yt.ProjectionPlot(sim_ds, "x", "density", center=(0.5,0.3,0.3)) p.annotate_halos(hc) ______________________________________________________________

1 0

Why individual plots have to be saved before being handed to eps_writer?
by Anni Järvenpää 04 Aug '15

04 Aug '15

Hi All, I have been updating the eps_writer to handle a mixture of profiles, slices and projections that can be displayed in a single plot (multiple panels). eps_writer with my edits can be reviewed at http://paste.yt-project.org/show/5773/. However, in order to get the plot axes to work correctly I need to save the individual plots as well before passing them to the eps_writer. Here is an example script: http://paste.yt-project.org/show/5772/. It works correctly as is, but uncommenting the changes in plots causes weird results if I don't also save the individual plots. The result of saving and not-saving are below (as you can see the plot doesn't change but the axes/colorbar does). http://www.mv.helsinki.fi/home/aajarven/yt/noedit.png http://www.mv.helsinki.fi/home/aajarven/yt/edits-nosave.png http://www.mv.helsinki.fi/home/aajarven/yt/edits-save.png Does anyone know why the plots must be saved before being handed to the eps_writer api? I'd like to push these changes to the eps_writer and could possibly raise the save as a bug? Let me know. Anni

4 4

multipanel plot resolution
by Shankar Agarwal 01 Aug '15

01 Aug '15

Hi, When making a typical 1x1 plot, I see that the final image has 800x800 pixels. I then used the following script to make a 2x2 multipanel plot, with each panel showing the projected density: http://yt-project.org/doc/cookbook/complex_plots.html#time-series-multipanel However, I see that each of the four panels is not 800x800. It is about 388x388. As such, some of the details in each panel is lost (which I can see if I plot the four panels separately on 1x1 plots). Is there an easy way to tell the above script that each panel be 800x800 pixel? By the way, I see there is a statement Npixels = 512 in http://yt-project.org/doc/cookbook/complex_plots.html#off-axis-projection-a… Cheers, Shankar

2 1