Mailman 3 May 2014 - yt-users

Boundary of the data container
by Junhwan Choi (최준환) 07 May '14

07 May '14

Hi yt users, I have one basic question about the membership of the data container. For example, I make a sphere data container and try to estimate the total enclosed mass of the sphere. However one of the refined cell (grid) is partially enclosed in the sphere. Does the yt add entire cell (grid) mass into the sphere mass or estimate the volume fraction of the cell mass and add only this partial mass? This question can be also applied for the density profile (r vs density). What happen one cell extends to multiple binning point? I would like to clarify my yt analysis. Thanks in advance, Junhwan

5 6

Re: [yt-users] Parallel Broadcasting in yt
by Claire Kopenhafer 07 May '14

07 May '14

mpi_bcast was what I was looking for! Thank you -Claire On Tue, May 6, 2014 at 2:05 PM, <yt-users-request(a)lists.spacepope.org>wrote: > Send yt-users mailing list submissions to > yt-users(a)lists.spacepope.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org > or, via email, send a message with subject or body 'help' to > yt-users-request(a)lists.spacepope.org > > You can reach the person managing the list at > yt-users-owner(a)lists.spacepope.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of yt-users digest..." > > Today's Topics: > > 1. Parallel Broadcasting in yt (Claire Kopenhafer) > 2. Re: Parallel Broadcasting in yt (Britton Smith) > 3. fractional keyword in profile plot (Chang-Goo Kim) > 4. total mass (Chang-Goo Kim) > 5. Re: total mass (Matthew Turk) > > > ---------- Forwarded message ---------- > From: Claire Kopenhafer <clairekope(a)gmail.com> > To: yt-users(a)lists.spacepope.org > Cc: > Date: Mon, 5 May 2014 15:56:29 -0400 > Subject: [yt-users] Parallel Broadcasting in yt > Hello all, > > I've got a code that reads in a dictionary from file and then operates on > it, and I'd like to run it in parallel. My best thought is to have the root > process read in the dict and then broadcast it to all the rest, but I am > not sure how best to do this within yt. Any suggestions? Thanks! > > -Claire > > > ---------- Forwarded message ---------- > From: Britton Smith <brittonsmith(a)gmail.com> > To: Discussion of the yt analysis package <yt-users(a)lists.spacepope.org> > Cc: > Date: Mon, 5 May 2014 16:50:06 -0400 > Subject: Re: [yt-users] Parallel Broadcasting in yt > Hi Claire, > > You might want to have a look at the function called mpi_bcast > in yt/utilities/parallel_tools/parallel_analysis_interface.py. This should > do a standard mpi broadcast from the root process to all of the > subprocesses. Alternatively, if the dictionary is something that can be > loaded in in parts by each of the processors and then combined, you might > also want to check par_combine_object in the same file. > > Britton > > > On Mon, May 5, 2014 at 3:56 PM, Claire Kopenhafer <clairekope(a)gmail.com>wrote: > >> Hello all, >> >> I've got a code that reads in a dictionary from file and then operates on >> it, and I'd like to run it in parallel. My best thought is to have the root >> process read in the dict and then broadcast it to all the rest, but I am >> not sure how best to do this within yt. Any suggestions? Thanks! >> >> -Claire >> >> _______________________________________________ >> yt-users mailing list >> yt-users(a)lists.spacepope.org >> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org >> >> > > > ---------- Forwarded message ---------- > From: Chang-Goo Kim <cgkim(a)astro.princeton.edu> > To: Discussion of the yt analysis package <yt-users(a)lists.spacepope.org> > Cc: > Date: Tue, 6 May 2014 13:43:39 -0400 > Subject: [yt-users] fractional keyword in profile plot > Hi all, > > I tried to draw PDFs using ProfilePlot with "fractional" keyword. But, it > returned unit incompatibility. I have looked at the code, and realized that > the fractional keyword makes field_data dimensionless, while it tried to > return data with units of original data. I think it must be easy to fix. > Error message is attached below. > > Thanks, > > Chang-Goo > > ---------------------------------------------------------------------------YTUnitConversionError Traceback (most recent call last)<ipython-input-137-18016774c3d0> in <module>()----> 1 pdf=ProfilePlot(dd,"temp",['cell_mass','cell_volume'],weight_field=None,fractional=True) 2 pdf.show() > /u/cgkim/Sources/yt-x86_64/src/yt-hg/yt/visualization/profile_plotter.pyc in __init__(self, data_source, x_field, y_fields, weight_field, n_bins, accumulation, fractional, label, plot_spec, profiles) 216 self.plot_spec = [self.plot_spec.copy() for p in self.profiles] 217 --> 218 self._setup_plots() 219 220 def save(self, name=None): > /u/cgkim/Sources/yt-x86_64/src/yt-hg/yt/visualization/profile_plotter.pyc in _setup_plots(self) 309 self.axes = AxesContainer(self.figures) 310 for i, profile in enumerate(self.profiles):--> 311 for field, field_data in profile.items(): 312 self.axes[field].plot(np.array(profile.x), 313 np.array(field_data), > /u/cgkim/Sources/yt-x86_64/src/yt-hg/yt/data_objects/profiles.pyc in items(self) 854 855 def items(self):--> 856 return [(k,self[k]) for k in self.field_data.keys()] 857 858 def __iter__(self): > /u/cgkim/Sources/yt-x86_64/src/yt-hg/yt/data_objects/profiles.pyc in __getitem__(self, field) 851 raise KeyError(field) 852 else:--> 853 return self.field_data[fname].in_units(self.field_units[fname]) 854 855 def items(self): > /u/cgkim/Sources/yt-x86_64/src/yt-hg/yt/units/yt_array.pyc in in_units(self, units) 358 359 """--> 360 new_units = self._unit_repr_check_same(units) 361 conversion_factor = self.units.get_conversion_factor(new_units) 362 > /u/cgkim/Sources/yt-x86_64/src/yt-hg/yt/units/yt_array.pyc in _unit_repr_check_same(self, units) 315 if not self.units.same_dimensions_as(units): 316 raise YTUnitConversionError(--> 317 self.units, self.units.dimensions, units, units.dimensions) 318 319 return units > YTUnitConversionError: Unit dimensionalities do not match. Tried to convert between dimensionless (dim 1) and code_length**3 (dim (length)**3). > > > > > > ---------- Forwarded message ---------- > From: Chang-Goo Kim <cgkim(a)astro.princeton.edu> > To: Discussion of the yt analysis package <yt-users(a)lists.spacepope.org> > Cc: > Date: Tue, 6 May 2014 13:50:47 -0400 > Subject: [yt-users] total mass > Hi all again, > > I tried to find mass fraction of each temperature component, but I got > inconsistent results as below: > > == > > dd=pf.all_data() > > > > hot=dd.cut_region(["obj['temperature'] >= 1.e4"]) > > wnm=dd.cut_region(["obj['temperature'] < 1.e4"]) > > > > print dd.quantities.total_mass() > > print wnm.quantities.total_mass()+hot.quantities.total_mass() > > > > 2.88677799841e+38 g > 2.91575727709e+38 g > > == > > I can't understand why those two are different. Have any idea? > > Chang-Goo > > > > ---------- Forwarded message ---------- > From: Matthew Turk <matthewturk(a)gmail.com> > To: Discussion of the yt analysis package <yt-users(a)lists.spacepope.org> > Cc: > Date: Tue, 6 May 2014 13:53:45 -0400 > Subject: Re: [yt-users] total mass > Hi Chang-Goo, > > On Tue, May 6, 2014 at 1:50 PM, Chang-Goo Kim <cgkim(a)astro.princeton.edu> > wrote: > > Hi all again, > > > > I tried to find mass fraction of each temperature component, but I got > > inconsistent results as below: > > > > == > > > > dd=pf.all_data() > > > > > > > > hot=dd.cut_region(["obj['temperature'] >= 1.e4"]) > > > > wnm=dd.cut_region(["obj['temperature'] < 1.e4"]) > > > > > > > > print dd.quantities.total_mass() > > > > print wnm.quantities.total_mass()+hot.quantities.total_mass() > > > > The total_mass includes particle fields, which will be incorrectly cut > by the cut_region. We're trying to figure out a way to make this > clearer, but you should get identical results if you use the > .total_quantity("cell_mass") function. This was the subject of a > recent bug report and discussion, as well: > > > https://bitbucket.org/yt_analysis/yt/issue/812/derived-quantities-on-cut_re… > > I thought we'd had a pull request discussion, too, but I can't find it > right now. > > -Matt > > > > > > > 2.88677799841e+38 g > > 2.91575727709e+38 g > > > > == > > > > I can't understand why those two are different. Have any idea? > > > > Chang-Goo > > > > > > _______________________________________________ > > yt-users mailing list > > yt-users(a)lists.spacepope.org > > http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org > > > > > _______________________________________________ > yt-users mailing list > yt-users(a)lists.spacepope.org > http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org > >

1 0

fractional keyword in profile plot
by Chang-Goo Kim 06 May '14

06 May '14

Hi all, I tried to draw PDFs using ProfilePlot with "fractional" keyword. But, it returned unit incompatibility. I have looked at the code, and realized that the fractional keyword makes field_data dimensionless, while it tried to return data with units of original data. I think it must be easy to fix. Error message is attached below. Thanks, Chang-Goo ---------------------------------------------------------------------------YTUnitConversionError Traceback (most recent call last)<ipython-input-137-18016774c3d0> in <module>()----> 1 pdf=ProfilePlot(dd,"temp",['cell_mass','cell_volume'],weight_field=None,fractional=True) 2 pdf.show() /u/cgkim/Sources/yt-x86_64/src/yt-hg/yt/visualization/profile_plotter.pyc in __init__(self, data_source, x_field, y_fields, weight_field, n_bins, accumulation, fractional, label, plot_spec, profiles) 216 self.plot_spec = [self.plot_spec.copy() for p in self.profiles] 217 --> 218 self._setup_plots() 219 220 def save(self, name=None): /u/cgkim/Sources/yt-x86_64/src/yt-hg/yt/visualization/profile_plotter.pyc in _setup_plots(self) 309 self.axes = AxesContainer(self.figures) 310 for i, profile in enumerate(self.profiles):--> 311 for field, field_data in profile.items(): 312 self.axes[field].plot(np.array(profile.x), 313 np.array(field_data), /u/cgkim/Sources/yt-x86_64/src/yt-hg/yt/data_objects/profiles.pyc in items(self) 854 855 def items(self):--> 856 return [(k,self[k]) for k in self.field_data.keys()] 857 858 def __iter__(self): /u/cgkim/Sources/yt-x86_64/src/yt-hg/yt/data_objects/profiles.pyc in __getitem__(self, field) 851 raise KeyError(field) 852 else:--> 853 return self.field_data[fname].in_units(self.field_units[fname]) 854 855 def items(self): /u/cgkim/Sources/yt-x86_64/src/yt-hg/yt/units/yt_array.pyc in in_units(self, units) 358 359 """--> 360 new_units = self._unit_repr_check_same(units) 361 conversion_factor = self.units.get_conversion_factor(new_units) 362 /u/cgkim/Sources/yt-x86_64/src/yt-hg/yt/units/yt_array.pyc in _unit_repr_check_same(self, units) 315 if not self.units.same_dimensions_as(units): 316 raise YTUnitConversionError(--> 317 self.units, self.units.dimensions, units, units.dimensions) 318 319 return units YTUnitConversionError: Unit dimensionalities do not match. Tried to convert between dimensionless (dim 1) and code_length**3 (dim (length)**3).

2 1

total mass
by Chang-Goo Kim 06 May '14

06 May '14

Hi all again, I tried to find mass fraction of each temperature component, but I got inconsistent results as below: == dd=pf.all_data() hot=dd.cut_region(["obj['temperature'] >= 1.e4"]) wnm=dd.cut_region(["obj['temperature'] < 1.e4"]) print dd.quantities.total_mass() print wnm.quantities.total_mass()+hot.quantities.total_mass() 2.88677799841e+38 g 2.91575727709e+38 g == I can't understand why those two are different. Have any idea? Chang-Goo

2 1

Parallel Broadcasting in yt
by Claire Kopenhafer 05 May '14

05 May '14

Hello all, I've got a code that reads in a dictionary from file and then operates on it, and I'd like to run it in parallel. My best thought is to have the root process read in the dict and then broadcast it to all the rest, but I am not sure how best to do this within yt. Any suggestions? Thanks! -Claire

2 1

Profile of a projected quantity?
by Nathan Goldbaum 05 May '14

05 May '14

Hi all, Is there a way for me to find a profile using a projection as a data source? In particular, I'd like to find a radial profile of a surface density in a disk galaxy simulation. I think I could do this by histogramming a fixed resolution buffer created by a projection data source but is that the only way to do it? I'd prefer for my data pipeline to look identical for computing profiles and in all other cases I'm using a Profile1D object to find things like the cumulative mass or the average density as a function of radius. Thanks for your help with this! -Nathan

5 7

Problem importing yt.mods
by Saqib Nawaz 05 May '14

05 May '14

hello everyone, i am new to enzo. I am tyrying to make a phase plot from a.py script. but i am getting following errors. from: can't read /var/mail/yt.mods ./a.py: line 4: syntax error near unexpected token `(' ./a.py: line 4: `pf = load("IsolatedGalaxy/galaxy0030")' Is yt is not activated right? yt_install.log file is also attached. thanks in advance Saqib

2 2

installation problem.
by oleg burgueño 05 May '14

05 May '14

i been having a problem trying to install the 3.0 version of yt. It appears that some file or files are missing. i would appreciate any help in the matter, thank you

2 1

Axis labels and text annotations
by Jeremy Ritter 04 May '14

04 May '14

Hello, I'm relatively new to YT. Creating great looking plots seems to be really easy, but setting axis labels and other text labels (timestamps) has proven to be quite challenging. I'm creating some phase plots with the latest stable version (2.6 I think), and I would like for variables to have proper names, e.g. "Number Density" instead of "NumberDensity". I first tried changing the labels using the axes object set_xlabel/set_ylabel methods, but that removes the units and doesn't accept latex style commands for making my own units. It also appears to change the font style. Then I tried creating derived fields with the proper names including spaces, but it seems that the labels get squished, having all spaces removed when it prints the label. Filename output still has the spaces in the variable name. What is the easiest way to set publication quality axis labels with units in v2.6? My other problem seems to be that I'm unable to use any of the annotation callbacks with a phase plot. Is this true? I just want to add a timestamp, for which I've already created the string, e.g. "5.35 Myr". Thanks! -Jeremy

2 2

Freeprocessing
by tom fogal 01 May '14

01 May '14

Hi all, I'm Tom; new here, but I've been lurking on #yt (and #enzo) for a while now. My group and I created an in situ visualization tool that we're calling "Freeprocessing", and a few of you helped me hook it up to yt for a publication. Thanks for all your help---that paper was recently accepted! :-) Now that the edits are in, I'd like to delve into the problem deeper and hopefully expand on the 'proof of concept' we had used in the paper. If you could give it a try, I'd love to get people's thoughts on future directions, or even just issues we could address. More info on the website: https://tfogal.github.io/freeprocessing/ I'm "tfogal" on freenode/#yt and would love to chat with some of you there, but email works if you're not the IRC type. Thanks! -tom

2 1