Hello hope you are doing well. I am trying to save some projection and
slice field data within a sphere by doing the following :
my_sphere = stuff.sphere([0.5,0.5,0.5],(214,"kpc"))
prj1 = yt.SlicePlot(stuff,'z',('gas','radius'),data_source=my_sphere)
rad = prj1.frb['gas','radius']
prj1.save()
I then save rad to a pickle file to read in by another code. The full
simulation domain is 800 kpc. When I load the pickle file and print out the
values I am getting many 0's. And this is also the same when I make a
projection of some field and save it in the same way to a pickle file.
My question is, even though I am limiting the data to within a sphere, does
rad also hold values that are not within that sphere and saving them as
0's? Is there a better way to go about this?
Thank you.
--
Edward N. Buie II,
*Ph.D. Candidate*
Astrophysics
School of Earth and Space Exploration (SESE)
Arizona State University
Hi
I would like to create a new filed, but I encounter one problem here.
The dataset is in Arepo format, and there is one field ('PartType0', 'GFM_Metals'), the dimension of this field is (2716294, 10). For normal field, like ('PartType0', 'Masses'), the dimension is (2716294, ). So field ('PartType0', 'GFM_Metals') has more dimensions.
Below is how I create the new field. As I only need one row from ('PartType0', 'GFM_Metals') , so that:
def _nOxy(field, data):
X_Oxygen = data['PartType0', 'GFM_Metals'][:,4]
rho = data['PartType0', 'Density']
nOxy = X_Oxygen *rho/ ds.quan(mh,'g')
return nOxy
ds.add_field(('PartType0', 'nOxy'), function=_nOxy,
units='cm**-3',particle_type=True)
However, it shows error:
IndexError: too many indices for array
The error is from data['PartType0', 'GFM_Metals'][:,4], where yt does not support using [:,4].
So I am wondering if there is any other method to handle this.
I am grateful for your suggestions.
Thanks,
Yang
Dear yt users and developers,
we would like to draw your attention on two upcoming papers of ours on
performance studies of yt. The first one will be on ArXiv on Friday
("Speeding simulation analysis up with yt and Intel Distribution for
Python", authors: S. Cielo, L. Iapichino and F. Baruffa) and shows the
performance gained when installing yt on top of a Conda environment
based on the Intel Distribution for Python, rather than on the
traditional Anaconda-based environment. The second paper is more general
in scope (code optimisation of astrophysical codes on Intel Xeon Phi
Knights Landing) but has a whole section on tests of different
parallelisation strategies in yt (including Cython and OpenMP). This
paper is still under review and we share a semi-final version with you
under this link:
https://syncandshare.lrz.de/dl/fiFjBeCCh63DWACCcf3cTjdm/?inline
The aim of both works is to explore the potential offered by the
software stack of modern HPC systems for applications like yt. Besides a
custom installation procedure, no modification of the source code was
required. We see these works as potential contributions to the yt user
community, and therefore we would welcome comments, discussions and
follow-up ideas about them. For those of you attending SC19 in November,
one of us (Salvatore) will be there and is available for face-to-face
meetings.
Looking forward to your feedback,
Best regards,
Luigi and Salvatore
--
-----------------------------------------
Dr. Luigi Iapichino
Leibniz-Rechenzentrum
High Performance Systems Division
Boltzmannstr. 1, 85748 Garching, Germany
Tel: +49 89 35831 7824
e-mail: luigi.iapichino(a)lrz.de
www: http://iapichino.userweb.mwn.de/
Hadoop is a member degree, open supply package framework made for storage and process of huge scale type of data on clusters of artifact hardware. The Apache Hadoop application library is a framework that allows the data distributed processing across clusters for computing using easy programming models called Map Reduce
https://www.sevenmentor.com/hadoop-admin-training-institute-pune.php
Hello yt users,
I would like to know how data is arranged on yt arrays given a data container. For example, if I have a disk data container
ds=yt.load(snap)
disk=ds.disk(center,normal,(radi,'pc'),(height,'pc'))
And I want to obtain the density:
dens=disk['dens']
Lets say that gives me a YT.array of shape (4000,). What I would like to know is how the 3D density is storage into this 1D array, for example if
the density is storage first along the 'x' axis, then the 'y' axis, and finally the 'z' axis, e.g. :
for i in range(x_dims):
for j in range(y_dims):
for k in range(z_dims):
disk['dens'] = density_inside_disk[i,j,k]
I don't know where to find that information in the yt source code. Do you know that? I'm not sure if the array disk.icoords provides that information.
Cheers,
I would like to annotate Mass weighted Magnetic field quivers on density projection plot so that i can display the magnetic field features.
Do annotate_magnetic_field() , annotate_quiver() work correctly on projection plot or do they work only on slice plot correctly.
If they work on projection plot then against which quantity magnetic field quivers are weighed or not weighted?
Hi all, I'm trying to use "gas", "radial velocity" to compare inflows and outflows in a simulation, but the results don't really look right.
using radial_velocity as a binary field, I defined new fields as follows
def inflow_density(field,data):
tr = data['gas','density']*((data['gas','radial_velocity'])<0).astype(float)
return tr
def outflow_density(field,data):
tr = data['gas','density']*((data['gas','radial_velocity'])>0).astype(float)
return tr
A ProjectionPlot gives one side which is all inflow, one side which is all outflow. This resembles what you get if you have an incorrect bulk_velocity field, but when I correct for it like this
sp = ds.sphere(center, (Rvir, "kpc"))
bulk_velocity = sp.quantities.bulk_velocity()
plot = yt.ProjectionPlot(ds, 0, center=center, fields=('gas','inflow_density'), width = (w,"kpc"), field_parameters={'bulk_velocity':bulk_velocity})
it doesn't change at all. Am I using this right?
Thanks,
Clayton
