3D field and spherical object
Hello! I want to create 3D field within a spherical object. For example, I create 3D electron density in the following manner: ds = yt.load("DATA") dd = ds.all_data() max_level = ds.index.max_level ref = 2**max_level low = ds.domain_left_edge dims = ds.domain_dimensions*ref nx, ny, nz = dims L = (ds.domain_right_edge - ds.domain_left_edge).d all_data_level_max = ds.covering_grid(level=max_level,left_edge=low,dims=dims) def _elec_den(field,data): N_A = 1/1.6726219e-24 X = 0.70 return ((all_data_level_max['gas','density'])*N_A*(1+X)/2/ds.quan(1,'g') ) ds.add_field(('gas','elec_den'), function=_elec_den, units='auto', dimensions=dimensions.number_density) Then if I do: dd['elec_den'] it will give me a 3D field which will have the same dimensions as the root grid of the dataset. However, if I create spheres and then a spherical object between these two spheres: rvir = 4.2454307068129566e+24 #virial radius of a halo [cm]. xyz = [0.4208984375, 0.2216796875, 0.5185546875] #centre of a halo. sph1 = ds.sphere(xyz, (2*rvir, 'cm')) sph2 = ds.sphere(xyz, (4*rvir, 'cm')) cutout_obj = sph2 - sph1 cutout_obj['elec_den'] or sph1['elec_den'] and sph2['elec_den'] will have the same dimensions as dd['elec_den'] ; but this cannot be true since sph1 and sph2 include less grid points. Is there a way I can create 3D field and use it within a different yt objects? Thanks, Salome
Hi! I think that there might be something strange going on with your field definition, but it's also entirely possible that I'm misinterpreting. The way you have it set up, I *think* that the results of the derived field (gas, elec_den) will always have the same shape as your all_data_level_max. I believe what you may want to do is change your field definition to use data['gas','density'] instead. Additionally, can you try: cutout_obj['index','ones'] and comparing the shape of that to the others? If that shape is correct then we can narrow it down to the way elec_den is defined. Thanks, Matt On Sun, Jan 31, 2021 at 8:41 AM Salome Mtchedlidze <salomchedlidze@gmail.com> wrote:
Hello!
I want to create 3D field within a spherical object. For example, I create 3D electron density in the following manner:
ds = yt.load("DATA") dd = ds.all_data()
max_level = ds.index.max_level ref = 2**max_level low = ds.domain_left_edge dims = ds.domain_dimensions*ref nx, ny, nz = dims L = (ds.domain_right_edge - ds.domain_left_edge).d all_data_level_max = ds.covering_grid(level=max_level,left_edge=low,dims=dims)
def _elec_den(field,data): N_A = 1/1.6726219e-24 X = 0.70 return ((all_data_level_max['gas','density'])*N_A*(1+X)/2/ds.quan(1,'g') ) ds.add_field(('gas','elec_den'), function=_elec_den, units='auto', dimensions=dimensions.number_density)
Then if I do: dd['elec_den'] it will give me a 3D field which will have the same dimensions as the root grid of the dataset. However, if I create spheres and then a spherical object between these two spheres:
rvir = 4.2454307068129566e+24 #virial radius of a halo [cm]. xyz = [0.4208984375, 0.2216796875, 0.5185546875] #centre of a halo. sph1 = ds.sphere(xyz, (2*rvir, 'cm')) sph2 = ds.sphere(xyz, (4*rvir, 'cm')) cutout_obj = sph2 - sph1
cutout_obj['elec_den'] or sph1['elec_den'] and sph2['elec_den'] will have the same dimensions as dd['elec_den'] ; but this cannot be true since sph1 and sph2 include less grid points. Is there a way I can create 3D field and use it within a different yt objects?
Thanks, Salome _______________________________________________ yt-users mailing list -- yt-users@python.org To unsubscribe send an email to yt-users-leave@python.org https://mail.python.org/mailman3/lists/yt-users.python.org/ Member address: matthewturk@gmail.com
Hello, Ah yes, you are right that will always give me the same shape for that field. The thing is that I have tried before what you said but then I don't get the field in 3D. For example: def _create_elec_den(ds): def _elec_den(field,data): N_A = 1/1.6726219e-24 X = 0.70 return ((data['gas','density'])*N_A*(1+X)/2/ds.quan(1,'g')) ds.add_field(('gas','elec_den'), function=_elec_den, units='1/cm**3', force_override=True) and then: _create_elec_den(ds) cutout_obj['elec_den'] will have a shape: (85852,) and not e.g.(44,44,44). So, probably there is a proper way of doing it.. Thank you, Salome
Hi Salome, Ah, I see what you mean. So this is a bit tricky; the ordering of cells inside a data object (like a sphere, or a cutout) is not well-defined, so we return them as flattened arrays. In particular, when you have any different resolution points collected within the cutout/sphere, it's not obvious how to return these in 3D. Is it sufficient to be able to get the x,y,z coordinates, in addition? Or, would it work to mask out the cells (with either 0 or NaN) in a 3D array that match up with the cutouts? On Sun, Jan 31, 2021 at 1:36 PM Salome Mtchedlidze <salomchedlidze@gmail.com> wrote:
Hello,
Ah yes, you are right that will always give me the same shape for that field. The thing is that I have tried before what you said but then I don't get the field in 3D. For example:
def _create_elec_den(ds): def _elec_den(field,data): N_A = 1/1.6726219e-24 X = 0.70 return ((data['gas','density'])*N_A*(1+X)/2/ds.quan(1,'g')) ds.add_field(('gas','elec_den'), function=_elec_den, units='1/cm**3', force_override=True)
and then: _create_elec_den(ds) cutout_obj['elec_den']
will have a shape: (85852,) and not e.g.(44,44,44). So, probably there is a proper way of doing it..
Thank you, Salome _______________________________________________ yt-users mailing list -- yt-users@python.org To unsubscribe send an email to yt-users-leave@python.org https://mail.python.org/mailman3/lists/yt-users.python.org/ Member address: matthewturk@gmail.com
Hi Salome, OK, so I *think* this should work, but I'm not 100%. Try creating your covering grid, then before you access any fields, do this: cg._data_source = cutout_region Then when you access it, it *should* only access items inside the cutout region. On Sun, Jan 31, 2021 at 1:57 PM Salome Mtchedlidze <salomchedlidze@gmail.com> wrote:
yes, masking out the cells in a 3D array should be sufficient!
thanks, Salome _______________________________________________ yt-users mailing list -- yt-users@python.org To unsubscribe send an email to yt-users-leave@python.org https://mail.python.org/mailman3/lists/yt-users.python.org/ Member address: matthewturk@gmail.com
Hi Matthew, Did you mean something like this? def _create_elec_den(ds): def _elec_den(field,data): ----- return () all_data_level_max = ds.covering_grid(----) cutout_obj = sph2 - sph1 all_data_level_max._data_source = cutout_obj _create_elec_den(ds) If it is so, then cutout_obj['elec_den'] gives me again flattened array..
Yup, just about -- but try instead doing all_data_level_max['elec_den']. I *think* it will be zero in the places not included in the cutout. On Sun, Jan 31, 2021 at 4:10 PM Salome Mtchedlidze <salomchedlidze@gmail.com> wrote:
Hi Matthew, Did you mean something like this?
def _create_elec_den(ds): def _elec_den(field,data): ----- return ()
all_data_level_max = ds.covering_grid(----) cutout_obj = sph2 - sph1 all_data_level_max._data_source = cutout_obj _create_elec_den(ds)
If it is so, then cutout_obj['elec_den'] gives me again flattened array.. _______________________________________________ yt-users mailing list -- yt-users@python.org To unsubscribe send an email to yt-users-leave@python.org https://mail.python.org/mailman3/lists/yt-users.python.org/ Member address: matthewturk@gmail.com
participants (2)
-
Matthew Turk -
Salome Mtchedlidze