Can XrayLightCone work for Illustris simulation?

[tekkamanmaverick]

Hi John, I want to use the data of Illustris to generate light cone simulations of X-rays. Could I use the XrayLightCone object to do this? If not, could you expand it to Illustris? Thank you.

[jzuhone]

@tekkamanmaverick sorry for the late response.

The XrayLightCone should work for Illustris--if you see any odd behavior or errors please report them.

[tekkamanmaverick]

As you write in your Web page, XrayLightCone only works for Enzo or Gadget-based cosmological simulations. However, Illustris is based on Arepo which is different with Enzo and Gadget. Also your tutorials don't mention Illustris. Could you expand it to Illustris? Thank you.

[jzuhone]

@tekkamanmaverick It does work on Illustris data (I've done it before)--the only problem is that pyXSIM wants a "smoothing length" field to determine the extent of the X-ray emission from each particle, which the Illustris TNG outputs do not have.

You could let yt define a smoothing length field for you. You would need code like this:

from yt.fields.particle_fields import add_nearest_neighbor_field

ds = yt.load("illustris_data.hdf5")

num_neighbors = 64
add_nearest_neighbor_field("PartType0", "particle_position", ds, num_neighbors)
def _smoothing_length(field, data):
    rv = data["PartType0", 'nearest_neighbor_distance_%d' % num_neighbors]
    return rv
ds.add_field(("PartType0", "smoothing_length"), sampling_type="particle",
            function=_smoothing_length, units="code_length")

# Then pyXSIM code happens afterward

[tekkamanmaverick]

According to your codes, XrayLightCone needs parameters for the simulation like 'Enzo_64/64.param' in Enzo which the Illustris outputs do not have. How did you make it ?

[jzuhone]

Hi @tekkamanmaverick, sorry for the delayed reply.

I realize now that this is a bit of a problem--one would need the original AREPO parameter file that the simulation was run with, which I don't think the Illustris folks have made available. I will make a note to see if we can find a way to make the light cones without needing to have that.

[tekkamanmaverick]

Thank you. I hope pyXSIM will be more powerful.

[tekkamanmaverick]

yt : [WARNING ] Field ('PartType0', 'smoothing_length') already exists.

[jzuhone]

@tekkamanmaverick, sorry, what's the context for this last comment? I've seen that warning message before but I'd need to know how you got there, i.e. what version of pyXSIM, what's your script, etc.

[tekkamanmaverick]

In [1]: import yt

In [2]: import pyxsim /work/share/software/miniconda2/lib/python2.7/site-packages/astropy/table/np_utils.py:19: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88 from . import _np_utils /work/share/software/miniconda2/lib/python2.7/site-packages/astropy/io/ascii/fastbasic.py:9: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88 from . import cparser /work/share/software/miniconda2/lib/python2.7/site-packages/yt/analysis_modules/cosmological_observation/api.py:22: VisibleDeprecationWarning: Development of the LightRay module has been moved to the Trident package. This version is deprecated and will be removed from yt in a future release. See https://github.com/trident-project/trident for further information. from .light_ray.api import \

In [3]: from yt.fields.particle_fields import add_nearest_neighbor_field

In [4]: ds = yt.load('/scratch/Illustris/Illustris-3/snapdir_131/sna ...: p_131.0.hdf5') yt : [INFO ] 2018-07-31 18:02:48,584 Calculating time from 9.537e-01 to be 4.134e+17 seconds yt : [INFO ] 2018-07-31 18:02:48,585 Assuming length units are in kpc/h (comoving) yt : [INFO ] 2018-07-31 18:02:48,692 Parameters: current_time = 4.133936904150802e+17 s yt : [INFO ] 2018-07-31 18:02:48,693 Parameters: domain_dimensions = [2 2 2] yt : [INFO ] 2018-07-31 18:02:48,694 Parameters: domain_left_edge = [0. 0. 0.] yt : [INFO ] 2018-07-31 18:02:48,695 Parameters: domain_right_edge = [75000. 75000. 75000.] yt : [INFO ] 2018-07-31 18:02:48,696 Parameters: cosmological_simulation = 1 yt : [INFO ] 2018-07-31 18:02:48,696 Parameters: current_redshift = 0.04852362998180593 yt : [INFO ] 2018-07-31 18:02:48,696 Parameters: omega_lambda = 0.7274 yt : [INFO ] 2018-07-31 18:02:48,696 Parameters: omega_matter = 0.2726 yt : [INFO ] 2018-07-31 18:02:48,697 Parameters: hubble_constant = 0.704

In [5]: num_neighbors = 64

In [6]: add_nearest_neighbor_field("PartType0", "particle_position", ds, num_nei ...: ghbors) yt : [INFO ] 2018-07-31 18:03:27,867 Allocating for 2.804e+08 particles (index particle type 'all') yt : [INFO ] 2018-07-31 18:04:38,162 Identified 1.316e+07 octs /work/zhangyuning/share/software/miniconda2/lib/python2.7/site-packages/yt/units/yt_array.py:978: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use arr[tuple(seq)] instead of arr[seq]. In the future this will be interpreted as an array index, arr[np.array(seq)], which will result either in an error or a different result. ret = super(YTArray, self).getitem(item) /work/zhangyuning/share/software/miniconda2/lib/python2.7/site-packages/yt/fields/fluid_fields.py:201: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use arr[tuple(seq)] instead of arr[seq]. In the future this will be interpreted as an array index, arr[np.array(seq)], which will result either in an error or a different result. new_field[slice_3d] = f Out[6]: [('PartType0', 'nearest_neighbor_distance_64')]

In [7]: def _smoothing_length(field, data): ...: rv = data["PartType0", 'nearest_neighbordistance%d' % num_neighbor ...: s] ...: return rv ...:

In [8]: ds.add_field(("PartType0", "smoothing_length"), sampling_type="particle" ...: ,function=_smoothing_length, units="code_length") yt : [WARNING ] 2018-07-31 18:05:37,248 Field ('PartType0', 'smoothing_length') already exists. To override use force_override=True.

The version is 2.1.1.